addyolo

 code/web/package-lock.json
 code/web/node_modules/**
+/code/venv
+/code/deep_sort_yolov4/model_data/yolov4.weights
+/code/deep_sort_yolov4/model_data/yolo4_weight.h5
\ No newline at end of file

+# YOLOv4 + Deep_SORT
+
+<img src="https://github.com/yehengchen/Object-Detection-and-Tracking/blob/master/OneStage/yolo/deep_sort_yolov4/output/comparison.png" width="81%" height="81%"> <img src="https://github.com/yehengchen/video_demo/blob/master/video_demo/output.gif" width="40%" height="40%"> <img src="https://github.com/yehengchen/video_demo/blob/master/video_demo/TownCentreXVID_output.gif" width="40%" height="40%">
+
+__Object Tracking & Counting Demo - [[BiliBili]](https://www.bilibili.com/video/BV1Ug4y1i71w#reply3014975828)  [[Chinese Version]](https://blog.csdn.net/weixin_38107271/article/details/96741706)__
+## Requirement
+__Development Environment: [Deep-Learning-Environment-Setup](https://github.com/yehengchen/Ubuntu-16.04-Deep-Learning-Environment-Setup)__ 
+
+* OpenCV
+* sklean
+* pillow
+* numpy 1.15.0
+* torch 1.3.0
+* tensorflow-gpu 1.13.1
+* CUDA 10.0
+***
+
+It uses:
+
+* __Detection__: [YOLOv4](https://github.com/yehengchen/Object-Detection-and-Tracking/tree/master/OneStage/yolo/Train-a-YOLOv4-model) to detect objects on each of the video frames. - 用自己的数据训练YOLOv4模型
+
+* __Tracking__: [Deep_SORT](https://github.com/nwojke/deep_sort) to track those objects over different frames.
+
+*This repository contains code for Simple Online and Realtime Tracking with a Deep Association Metric (Deep SORT). We extend the original SORT algorithm to integrate appearance information based on a deep appearance descriptor. See the [arXiv preprint](https://arxiv.org/abs/1703.07402) for more information.*
+
+## Quick Start
+
+__0.Requirements__
+
+    pip install -r requirements.txt
+    
+__1. Download the code to your computer.__
+    
+    git clone https://github.com/yehengchen/Object-Detection-and-Tracking.git
+    
+__2. Download [[yolov4.weights]](https://drive.google.com/file/d/1cewMfusmPjYWbrnuJRuKhPMwRe_b9PaT/view) [[Baidu]](https://pan.baidu.com/s/1jRudrrXAS3DRGqT6mL4L3A ) - `mnv6`__ and place it in `deep_sort_yolov4/model_data/`
+
+*Here you can download my trained [[yolo4_weight.h5]](https://pan.baidu.com/s/1JuT4KCUFaE2Gvme0_S37DQ ) - `w17w` weights for detecting person/car/bicycle,etc.*
+
+__3. Convert the Darknet YOLO model to a Keras model:__
+```
+$ python convert.py model_data/yolov4.cfg model_data/yolov4.weights model_data/yolo.h5
+``` 
+__4. Run the YOLO_DEEP_SORT:__
+
+```
+$ python main.py -c [CLASS NAME] -i [INPUT VIDEO PATH]
+
+$ python main.py -c person -i ./test_video/testvideo.avi
+```
+
+__5. Can change [deep_sort_yolov3/yolo.py] `__Line 100__` to your tracking object__
+
+*DeepSORT pre-trained weights using people-ReID datasets only for person*
+```
+    if predicted_class != args["class"]:
+               continue
+    
+    if predicted_class != 'person' and predicted_class != 'car':
+               continue
+```
+
+## Train on Market1501 & MARS
+*People Re-identification model*
+
+[cosine_metric_learning](https://github.com/nwojke/cosine_metric_learning) for training a metric feature representation to be used with the deep_sort tracker.
+
+## Citation
+
+### YOLOv4 :
+
+    @misc{bochkovskiy2020yolov4,
+    title={YOLOv4: Optimal Speed and Accuracy of Object Detection},
+    author={Alexey Bochkovskiy and Chien-Yao Wang and Hong-Yuan Mark Liao},
+    year={2020},
+    eprint={2004.10934},
+    archivePrefix={arXiv},
+    primaryClass={cs.CV}
+    }
+
+### Deep_SORT :
+
+    @inproceedings{Wojke2017simple,
+    title={Simple Online and Realtime Tracking with a Deep Association Metric},
+    author={Wojke, Nicolai and Bewley, Alex and Paulus, Dietrich},
+    booktitle={2017 IEEE International Conference on Image Processing (ICIP)},
+    year={2017},
+    pages={3645--3649},
+    organization={IEEE},
+    doi={10.1109/ICIP.2017.8296962}
+    }
+
+    @inproceedings{Wojke2018deep,
+    title={Deep Cosine Metric Learning for Person Re-identification},
+    author={Wojke, Nicolai and Bewley, Alex},
+    booktitle={2018 IEEE Winter Conference on Applications of Computer Vision (WACV)},
+    year={2018},
+    pages={748--756},
+    organization={IEEE},
+    doi={10.1109/WACV.2018.00087}
+    }
+    
+## Reference
+#### Github:deep_sort@[Nicolai Wojke nwojke](https://github.com/nwojke/deep_sort)
+#### Github:deep_sort_yolov3@[Qidian213 ](https://github.com/Qidian213/deep_sort_yolov3)
+#### Github:Deep-SORT-YOLOv4@[LeonLok](https://github.com/LeonLok/Deep-SORT-YOLOv4)
+

+import os
+import colorsys
+
+import numpy as np
+from keras import backend as K
+from keras.models import load_model
+from keras.layers import Input
+
+from yolo4.model import yolo_eval, yolo4_body
+from yolo4.utils import letterbox_image
+
+from PIL import Image, ImageFont, ImageDraw
+from timeit import default_timer as timer
+import matplotlib.pyplot as plt
+
+from operator import itemgetter
+
+class Yolo4(object):
+    def get_class(self):
+        classes_path = os.path.expanduser(self.classes_path)
+        with open(classes_path) as f:
+            class_names = f.readlines()
+        class_names = [c.strip() for c in class_names]
+        return class_names
+
+    def get_anchors(self):
+        anchors_path = os.path.expanduser(self.anchors_path)
+        with open(anchors_path) as f:
+            anchors = f.readline()
+        anchors = [float(x) for x in anchors.split(',')]
+        return np.array(anchors).reshape(-1, 2)
+
+    def load_yolo(self):
+        model_path = os.path.expanduser(self.model_path)
+        assert model_path.endswith('.h5'), 'Keras model or weights must be a .h5 file.'
+
+        self.class_names = self.get_class()
+        self.anchors = self.get_anchors()
+
+        num_anchors = len(self.anchors)
+        num_classes = len(self.class_names)
+
+        # Generate colors for drawing bounding boxes.
+        hsv_tuples = [(x / len(self.class_names), 1., 1.)
+                      for x in range(len(self.class_names))]
+        self.colors = list(map(lambda x: colorsys.hsv_to_rgb(*x), hsv_tuples))
+        self.colors = list(
+            map(lambda x: (int(x[0] * 255), int(x[1] * 255), int(x[2] * 255)),
+                self.colors))
+
+        self.sess = K.get_session()
+
+        # Load model, or construct model and load weights.
+        self.yolo4_model = yolo4_body(Input(shape=(608, 608, 3)), num_anchors//3, num_classes)
+
+        # Read and convert darknet weight
+        print('Loading weights.')
+        weights_file = open(self.weights_path, 'rb')
+        major, minor, revision = np.ndarray(
+            shape=(3, ), dtype='int32', buffer=weights_file.read(12))
+        if (major*10+minor)>=2 and major<1000 and minor<1000:
+            seen = np.ndarray(shape=(1,), dtype='int64', buffer=weights_file.read(8))
+        else:
+            seen = np.ndarray(shape=(1,), dtype='int32', buffer=weights_file.read(4))
+        print('Weights Header: ', major, minor, revision, seen)
+
+        convs_to_load = []
+        bns_to_load = []
+        for i in range(len(self.yolo4_model.layers)):
+            layer_name = self.yolo4_model.layers[i].name
+            if layer_name.startswith('conv2d_'):
+                convs_to_load.append((int(layer_name[7:]), i))
+            if layer_name.startswith('batch_normalization_'):
+                bns_to_load.append((int(layer_name[20:]), i))
+
+        convs_sorted = sorted(convs_to_load, key=itemgetter(0))
+        bns_sorted = sorted(bns_to_load, key=itemgetter(0))
+
+        bn_index = 0
+        for i in range(len(convs_sorted)):
+            print('Converting ', i)
+            if i == 93 or i == 101 or i == 109:
+                #no bn, with bias
+                weights_shape = self.yolo4_model.layers[convs_sorted[i][1]].get_weights()[0].shape
+                bias_shape = self.yolo4_model.layers[convs_sorted[i][1]].get_weights()[0].shape[3]
+                filters = bias_shape
+                size = weights_shape[0]
+                darknet_w_shape = (filters, weights_shape[2], size, size)
+                weights_size = np.product(weights_shape)
+
+                conv_bias = np.ndarray(
+                    shape=(filters, ),
+                    dtype='float32',
+                    buffer=weights_file.read(filters * 4))
+                conv_weights = np.ndarray(
+                    shape=darknet_w_shape,
+                    dtype='float32',
+                    buffer=weights_file.read(weights_size * 4))
+                conv_weights = np.transpose(conv_weights, [2, 3, 1, 0])
+                self.yolo4_model.layers[convs_sorted[i][1]].set_weights([conv_weights, conv_bias])
+            else:
+                #with bn, no bias
+                weights_shape = self.yolo4_model.layers[convs_sorted[i][1]].get_weights()[0].shape
+                size = weights_shape[0]
+                bn_shape = self.yolo4_model.layers[bns_sorted[bn_index][1]].get_weights()[0].shape
+                filters = bn_shape[0]
+                darknet_w_shape = (filters, weights_shape[2], size, size)
+                weights_size = np.product(weights_shape)
+
+                conv_bias = np.ndarray(
+                    shape=(filters, ),
+                    dtype='float32',
+                    buffer=weights_file.read(filters * 4))
+                bn_weights = np.ndarray(
+                    shape=(3, filters),
+                    dtype='float32',
+                    buffer=weights_file.read(filters * 12))
+
+                bn_weight_list = [
+                    bn_weights[0],  # scale gamma
+                    conv_bias,  # shift beta
+                    bn_weights[1],  # running mean
+                    bn_weights[2]  # running var
+                ]
+                self.yolo4_model.layers[bns_sorted[bn_index][1]].set_weights(bn_weight_list)
+
+                conv_weights = np.ndarray(
+                    shape=darknet_w_shape,
+                    dtype='float32',
+                    buffer=weights_file.read(weights_size * 4))
+                conv_weights = np.transpose(conv_weights, [2, 3, 1, 0])
+                self.yolo4_model.layers[convs_sorted[i][1]].set_weights([conv_weights])
+
+                bn_index += 1
+
+        weights_file.close()
+
+        self.yolo4_model.save(self.model_path)
+
+
+        if self.gpu_num>=2:
+            self.yolo4_model = multi_gpu_model(self.yolo4_model, gpus=self.gpu_num)
+
+        self.input_image_shape = K.placeholder(shape=(2, ))
+        self.boxes, self.scores, self.classes = yolo_eval(self.yolo4_model.output, self.anchors,
+                len(self.class_names), self.input_image_shape,
+                score_threshold=self.score)
+
+    def __init__(self, score, iou, anchors_path, classes_path, model_path, weights_path, gpu_num=1):
+        self.score = score
+        self.iou = iou
+        self.anchors_path = anchors_path
+        self.classes_path = classes_path
+        self.weights_path = weights_path
+        self.model_path = model_path
+        self.gpu_num = gpu_num
+        self.load_yolo()
+
+    def close_session(self):
+        self.sess.close()
+
+if __name__ == '__main__':
+    model_path = 'model_data/yolo4_weight.h5'
+    anchors_path = 'model_data/yolo_anchors.txt'
+    classes_path = 'model_data/coco_classes.txt'
+    weights_path = 'model_data/yolov4.weights'
+
+    score = 0.5
+    iou = 0.5
+
+    model_image_size = (608, 608)
+
+    yolo4_model = Yolo4(score, iou, anchors_path, classes_path, model_path, weights_path)
+
+    yolo4_model.close_session()
\ No newline at end of file

+# vim: expandtab:ts=4:sw=4

+# vim: expandtab:ts=4:sw=4
+import numpy as np
+
+
+class Detection(object):
+    """
+    This class represents a bounding box detection in a single image.
+
+    Parameters
+    ----------
+    tlwh : array_like
+        Bounding box in format `(x, y, w, h)`.
+    confidence : float
+        Detector confidence score.
+    feature : array_like
+        A feature vector that describes the object contained in this image.
+
+    Attributes
+    ----------
+    tlwh : ndarray
+        Bounding box in format `(top left x, top left y, width, height)`.
+    confidence : ndarray
+        Detector confidence score.
+    feature : ndarray | NoneType
+        A feature vector that describes the object contained in this image.
+
+    """
+
+    def __init__(self, tlwh, confidence, feature):
+        self.tlwh = np.asarray(tlwh, dtype=np.float)
+        self.confidence = float(confidence)
+        self.feature = np.asarray(feature, dtype=np.float32)
+
+    def to_tlbr(self):
+        """Convert bounding box to format `(min x, min y, max x, max y)`, i.e.,
+        `(top left, bottom right)`.
+        """
+        ret = self.tlwh.copy()
+        ret[2:] += ret[:2]
+        return ret
+
+    def to_xyah(self):
+        """Convert bounding box to format `(center x, center y, aspect ratio,
+        height)`, where the aspect ratio is `width / height`.
+        """
+        ret = self.tlwh.copy()
+        ret[:2] += ret[2:] / 2
+        ret[2] /= ret[3]
+        return ret

+# vim: expandtab:ts=4:sw=4
+import numpy as np
+
+
+class Detection_YOLO(object):
+    """
+    This class represents a bounding box detection in a single image.
+    Parameters
+    ----------
+    tlwh : array_like
+        Bounding box in format `(x, y, w, h)`.
+    confidence : float
+        Detector confidence score.
+    feature : array_like
+        A feature vector that describes the object contained in this image.
+    Attributesutils
+    ----------
+    tlwh : ndarray
+        Bounding box in format `(top left x, top left y, width, height)`.
+    confidence : ndarray
+        Detector confidence score.
+    feature : ndarray | NoneType
+        A feature vector that describes the object contained in this image.
+    """
+
+    def __init__(self, tlwh, confidence, cls):
+        self.tlwh = np.asarray(tlwh, dtype=np.float)
+        self.confidence = float(confidence)
+        self.cls = cls
+
+    def to_tlbr(self):
+        """Convert bounding box to format `(min x, min y, max x, max y)`, i.e.,
+        `(top left, bottom right)`.
+        """
+        ret = self.tlwh.copy()
+        ret[2:] += ret[:2]
+        return ret
+
+    def to_xyah(self):
+        """Convert bounding box to format `(center x, center y, aspect ratio,
+        height)`, where the aspect ratio is `width / height`.
+        """
+        ret = self.tlwh.copy()
+        ret[:2] += ret[2:] / 2
+        ret[2] /= ret[3]
+        return ret

+# vim: expandtab:ts=4:sw=4
+from __future__ import absolute_import
+import numpy as np
+from . import linear_assignment
+
+
+def iou(bbox, candidates):
+    """Computer intersection over union.
+
+    Parameters
+    ----------
+    bbox : ndarray
+        A bounding box in format `(top left x, top left y, width, height)`.
+    candidates : ndarray
+        A matrix of candidate bounding boxes (one per row) in the same format
+        as `bbox`.
+
+    Returns
+    -------
+    ndarray
+        The intersection over union in [0, 1] between the `bbox` and each
+        candidate. A higher score means a larger fraction of the `bbox` is
+        occluded by the candidate.
+
+    """
+    bbox_tl, bbox_br = bbox[:2], bbox[:2] + bbox[2:]
+    candidates_tl = candidates[:, :2]
+    candidates_br = candidates[:, :2] + candidates[:, 2:]
+
+    tl = np.c_[np.maximum(bbox_tl[0], candidates_tl[:, 0])[:, np.newaxis],
+               np.maximum(bbox_tl[1], candidates_tl[:, 1])[:, np.newaxis]]
+    br = np.c_[np.minimum(bbox_br[0], candidates_br[:, 0])[:, np.newaxis],
+               np.minimum(bbox_br[1], candidates_br[:, 1])[:, np.newaxis]]
+    wh = np.maximum(0., br - tl)
+
+    area_intersection = wh.prod(axis=1)
+    area_bbox = bbox[2:].prod()
+    area_candidates = candidates[:, 2:].prod(axis=1)
+    return area_intersection / (area_bbox + area_candidates - area_intersection)
+
+
+def iou_cost(tracks, detections, track_indices=None,
+             detection_indices=None):
+    """An intersection over union distance metric.
+
+    Parameters
+    ----------
+    tracks : List[deep_sort.track.Track]
+        A list of tracks.
+    detections : List[deep_sort.detection.Detection]
+        A list of detections.
+    track_indices : Optional[List[int]]
+        A list of indices to tracks that should be matched. Defaults to
+        all `tracks`.
+    detection_indices : Optional[List[int]]
+        A list of indices to detections that should be matched. Defaults
+        to all `detections`.
+
+    Returns
+    -------
+    ndarray
+        Returns a cost matrix of shape
+        len(track_indices), len(detection_indices) where entry (i, j) is
+        `1 - iou(tracks[track_indices[i]], detections[detection_indices[j]])`.
+
+    """
+    if track_indices is None:
+        track_indices = np.arange(len(tracks))
+    if detection_indices is None:
+        detection_indices = np.arange(len(detections))
+
+    cost_matrix = np.zeros((len(track_indices), len(detection_indices)))
+    for row, track_idx in enumerate(track_indices):
+        if tracks[track_idx].time_since_update > 1:
+            cost_matrix[row, :] = linear_assignment.INFTY_COST
+            continue
+
+        bbox = tracks[track_idx].to_tlwh()
+        candidates = np.asarray([detections[i].tlwh for i in detection_indices])
+        cost_matrix[row, :] = 1. - iou(bbox, candidates)
+    return cost_matrix

+# vim: expandtab:ts=4:sw=4
+import numpy as np
+import scipy.linalg
+
+
+"""
+Table for the 0.95 quantile of the chi-square distribution with N degrees of
+freedom (contains values for N=1, ..., 9). Taken from MATLAB/Octave's chi2inv
+function and used as Mahalanobis gating threshold.
+"""
+chi2inv95 = {
+    1: 3.8415,
+    2: 5.9915,
+    3: 7.8147,
+    4: 9.4877,
+    5: 11.070,
+    6: 12.592,
+    7: 14.067,
+    8: 15.507,
+    9: 16.919}
+
+
+class KalmanFilter(object):
+    """
+    A simple Kalman filter for tracking bounding boxes in image space.
+
+    The 8-dimensional state space
+
+        x, y, a, h, vx, vy, va, vh
+
+    contains the bounding box center position (x, y), aspect ratio a, height h,
+    and their respective velocities.
+
+    Object motion follows a constant velocity model. The bounding box location
+    (x, y, a, h) is taken as direct observation of the state space (linear
+    observation model).
+
+    """
+
+    def __init__(self):
+        ndim, dt = 4, 1.
+
+        # Create Kalman filter model matrices.
+        self._motion_mat = np.eye(2 * ndim, 2 * ndim)
+        for i in range(ndim):
+            self._motion_mat[i, ndim + i] = dt
+        self._update_mat = np.eye(ndim, 2 * ndim)
+
+        # Motion and observation uncertainty are chosen relative to the current
+        # state estimate. These weights control the amount of uncertainty in
+        # the model. This is a bit hacky.
+        self._std_weight_position = 1. / 20
+        self._std_weight_velocity = 1. / 160
+
+    def initiate(self, measurement):
+        """Create track from unassociated measurement.
+
+        Parameters
+        ----------
+        measurement : ndarray
+            Bounding box coordinates (x, y, a, h) with center position (x, y),
+            aspect ratio a, and height h.
+
+        Returns
+        -------
+        (ndarray, ndarray)
+            Returns the mean vector (8 dimensional) and covariance matrix (8x8
+            dimensional) of the new track. Unobserved velocities are initialized
+            to 0 mean.
+
+        """
+        mean_pos = measurement
+        mean_vel = np.zeros_like(mean_pos)
+        mean = np.r_[mean_pos, mean_vel]
+
+        std = [
+            2 * self._std_weight_position * measurement[3],
+            2 * self._std_weight_position * measurement[3],
+            1e-2,
+            2 * self._std_weight_position * measurement[3],
+            10 * self._std_weight_velocity * measurement[3],
+            10 * self._std_weight_velocity * measurement[3],
+            1e-5,
+            10 * self._std_weight_velocity * measurement[3]]
+        covariance = np.diag(np.square(std))
+        return mean, covariance
+
+    def predict(self, mean, covariance):
+        """Run Kalman filter prediction step.
+
+        Parameters
+        ----------
+        mean : ndarray
+            The 8 dimensional mean vector of the object state at the previous
+            time step.
+        covariance : ndarray
+            The 8x8 dimensional covariance matrix of the object state at the
+            previous time step.
+
+        Returns
+        -------
+        (ndarray, ndarray)
+            Returns the mean vector and covariance matrix of the predicted
+            state. Unobserved velocities are initialized to 0 mean.
+
+        """
+        std_pos = [
+            self._std_weight_position * mean[3],
+            self._std_weight_position * mean[3],
+            1e-2,
+            self._std_weight_position * mean[3]]
+        std_vel = [
+            self._std_weight_velocity * mean[3],
+            self._std_weight_velocity * mean[3],
+            1e-5,
+            self._std_weight_velocity * mean[3]]
+        motion_cov = np.diag(np.square(np.r_[std_pos, std_vel]))
+
+        mean = np.dot(self._motion_mat, mean)
+        covariance = np.linalg.multi_dot((
+            self._motion_mat, covariance, self._motion_mat.T)) + motion_cov
+
+        return mean, covariance
+
+    def project(self, mean, covariance):
+        """Project state distribution to measurement space.
+
+        Parameters
+        ----------
+        mean : ndarray
+            The state's mean vector (8 dimensional array).
+        covariance : ndarray
+            The state's covariance matrix (8x8 dimensional).
+
+        Returns
+        -------
+        (ndarray, ndarray)
+            Returns the projected mean and covariance matrix of the given state
+            estimate.
+
+        """
+        std = [
+            self._std_weight_position * mean[3],
+            self._std_weight_position * mean[3],
+            1e-1,
+            self._std_weight_position * mean[3]]
+        innovation_cov = np.diag(np.square(std))
+
+        mean = np.dot(self._update_mat, mean)
+        covariance = np.linalg.multi_dot((
+            self._update_mat, covariance, self._update_mat.T))
+        return mean, covariance + innovation_cov
+
+    def update(self, mean, covariance, measurement):
+        """Run Kalman filter correction step.
+
+        Parameters
+        ----------
+        mean : ndarray
+            The predicted state's mean vector (8 dimensional).
+        covariance : ndarray
+            The state's covariance matrix (8x8 dimensional).
+        measurement : ndarray
+            The 4 dimensional measurement vector (x, y, a, h), where (x, y)
+            is the center position, a the aspect ratio, and h the height of the
+            bounding box.
+
+        Returns
+        -------
+        (ndarray, ndarray)
+            Returns the measurement-corrected state distribution.
+
+        """
+        projected_mean, projected_cov = self.project(mean, covariance)
+
+        chol_factor, lower = scipy.linalg.cho_factor(
+            projected_cov, lower=True, check_finite=False)
+        kalman_gain = scipy.linalg.cho_solve(
+            (chol_factor, lower), np.dot(covariance, self._update_mat.T).T,
+            check_finite=False).T
+        innovation = measurement - projected_mean
+
+        new_mean = mean + np.dot(innovation, kalman_gain.T)
+        new_covariance = covariance - np.linalg.multi_dot((
+            kalman_gain, projected_cov, kalman_gain.T))
+        return new_mean, new_covariance
+
+    def gating_distance(self, mean, covariance, measurements,
+                        only_position=False):
+        """Compute gating distance between state distribution and measurements.
+
+        A suitable distance threshold can be obtained from `chi2inv95`. If
+        `only_position` is False, the chi-square distribution has 4 degrees of
+        freedom, otherwise 2.
+
+        Parameters
+        ----------
+        mean : ndarray
+            Mean vector over the state distribution (8 dimensional).
+        covariance : ndarray
+            Covariance of the state distribution (8x8 dimensional).
+        measurements : ndarray
+            An Nx4 dimensional matrix of N measurements, each in
+            format (x, y, a, h) where (x, y) is the bounding box center
+            position, a the aspect ratio, and h the height.
+        only_position : Optional[bool]
+            If True, distance computation is done with respect to the bounding
+            box center position only.
+
+        Returns
+        -------
+        ndarray
+            Returns an array of length N, where the i-th element contains the
+            squared Mahalanobis distance between (mean, covariance) and
+            `measurements[i]`.
+
+        """
+        mean, covariance = self.project(mean, covariance)
+        if only_position:
+            mean, covariance = mean[:2], covariance[:2, :2]
+            measurements = measurements[:, :2]
+
+        cholesky_factor = np.linalg.cholesky(covariance)
+        d = measurements - mean
+        z = scipy.linalg.solve_triangular(
+            cholesky_factor, d.T, lower=True, check_finite=False,
+            overwrite_b=True)
+        squared_maha = np.sum(z * z, axis=0)
+        return squared_maha

+# vim: expandtab:ts=4:sw=4
+from __future__ import absolute_import
+import numpy as np
+from sklearn.utils.linear_assignment_ import linear_assignment
+from . import kalman_filter
+
+
+INFTY_COST = 1e+5
+
+
+def min_cost_matching(
+        distance_metric, max_distance, tracks, detections, track_indices=None,
+        detection_indices=None):
+    """Solve linear assignment problem.
+
+    Parameters
+    ----------
+    distance_metric : Callable[List[Track], List[Detection], List[int], List[int]) -> ndarray
+        The distance metric is given a list of tracks and detections as well as
+        a list of N track indices and M detection indices. The metric should
+        return the NxM dimensional cost matrix, where element (i, j) is the
+        association cost between the i-th track in the given track indices and
+        the j-th detection in the given detection_indices.
+    max_distance : float
+        Gating threshold. Associations with cost larger than this value are
+        disregarded.
+    tracks : List[track.Track]
+        A list of predicted tracks at the current time step.
+    detections : List[detection.Detection]
+        A list of detections at the current time step.
+    track_indices : List[int]
+        List of track indices that maps rows in `cost_matrix` to tracks in
+        `tracks` (see description above).
+    detection_indices : List[int]
+        List of detection indices that maps columns in `cost_matrix` to
+        detections in `detections` (see description above).
+
+    Returns
+    -------
+    (List[(int, int)], List[int], List[int])
+        Returns a tuple with the following three entries:
+        * A list of matched track and detection indices.
+        * A list of unmatched track indices.
+        * A list of unmatched detection indices.
+
+    """
+    if track_indices is None:
+        track_indices = np.arange(len(tracks))
+    if detection_indices is None:
+        detection_indices = np.arange(len(detections))
+
+    if len(detection_indices) == 0 or len(track_indices) == 0:
+        return [], track_indices, detection_indices  # Nothing to match.
+
+    cost_matrix = distance_metric(
+        tracks, detections, track_indices, detection_indices)
+    cost_matrix[cost_matrix > max_distance] = max_distance + 1e-5
+    indices = linear_assignment(cost_matrix)
+
+    matches, unmatched_tracks, unmatched_detections = [], [], []
+    for col, detection_idx in enumerate(detection_indices):
+        if col not in indices[:, 1]:
+            unmatched_detections.append(detection_idx)
+    for row, track_idx in enumerate(track_indices):
+        if row not in indices[:, 0]:
+            unmatched_tracks.append(track_idx)
+    for row, col in indices:
+        track_idx = track_indices[row]
+        detection_idx = detection_indices[col]
+        if cost_matrix[row, col] > max_distance:
+            unmatched_tracks.append(track_idx)
+            unmatched_detections.append(detection_idx)
+        else:
+            matches.append((track_idx, detection_idx))
+    return matches, unmatched_tracks, unmatched_detections
+
+
+def matching_cascade(
+        distance_metric, max_distance, cascade_depth, tracks, detections,
+        track_indices=None, detection_indices=None):
+    """Run matching cascade.
+
+    Parameters
+    ----------
+    distance_metric : Callable[List[Track], List[Detection], List[int], List[int]) -> ndarray
+        The distance metric is given a list of tracks and detections as well as
+        a list of N track indices and M detection indices. The metric should
+        return the NxM dimensional cost matrix, where element (i, j) is the
+        association cost between the i-th track in the given track indices and
+        the j-th detection in the given detection indices.
+    max_distance : float
+        Gating threshold. Associations with cost larger than this value are
+        disregarded.
+    cascade_depth: int
+        The cascade depth, should be se to the maximum track age.
+    tracks : List[track.Track]
+        A list of predicted tracks at the current time step.
+    detections : List[detection.Detection]
+        A list of detections at the current time step.
+    track_indices : Optional[List[int]]
+        List of track indices that maps rows in `cost_matrix` to tracks in
+        `tracks` (see description above). Defaults to all tracks.
+    detection_indices : Optional[List[int]]
+        List of detection indices that maps columns in `cost_matrix` to
+        detections in `detections` (see description above). Defaults to all
+        detections.
+
+    Returns
+    -------
+    (List[(int, int)], List[int], List[int])
+        Returns a tuple with the following three entries:
+        * A list of matched track and detection indices.
+        * A list of unmatched track indices.
+        * A list of unmatched detection indices.
+
+    """
+    if track_indices is None:
+        track_indices = list(range(len(tracks)))
+    if detection_indices is None:
+        detection_indices = list(range(len(detections)))
+
+    unmatched_detections = detection_indices
+    matches = []
+    for level in range(cascade_depth):
+        if len(unmatched_detections) == 0:  # No detections left
+            break
+
+        track_indices_l = [
+            k for k in track_indices
+            if tracks[k].time_since_update == 1 + level
+        ]
+        if len(track_indices_l) == 0:  # Nothing to match at this level
+            continue
+
+        matches_l, _, unmatched_detections = \
+            min_cost_matching(
+                distance_metric, max_distance, tracks, detections,
+                track_indices_l, unmatched_detections)
+        matches += matches_l
+    unmatched_tracks = list(set(track_indices) - set(k for k, _ in matches))
+    return matches, unmatched_tracks, unmatched_detections
+
+
+def gate_cost_matrix(
+        kf, cost_matrix, tracks, detections, track_indices, detection_indices,
+        gated_cost=INFTY_COST, only_position=False):
+    """Invalidate infeasible entries in cost matrix based on the state
+    distributions obtained by Kalman filtering.
+
+    Parameters
+    ----------
+    kf : The Kalman filter.
+    cost_matrix : ndarray
+        The NxM dimensional cost matrix, where N is the number of track indices
+        and M is the number of detection indices, such that entry (i, j) is the
+        association cost between `tracks[track_indices[i]]` and
+        `detections[detection_indices[j]]`.
+    tracks : List[track.Track]
+        A list of predicted tracks at the current time step.
+    detections : List[detection.Detection]
+        A list of detections at the current time step.
+    track_indices : List[int]
+        List of track indices that maps rows in `cost_matrix` to tracks in
+        `tracks` (see description above).
+    detection_indices : List[int]
+        List of detection indices that maps columns in `cost_matrix` to
+        detections in `detections` (see description above).
+    gated_cost : Optional[float]
+        Entries in the cost matrix corresponding to infeasible associations are
+        set this value. Defaults to a very large value.
+    only_position : Optional[bool]
+        If True, only the x, y position of the state distribution is considered
+        during gating. Defaults to False.
+
+    Returns
+    -------
+    ndarray
+        Returns the modified cost matrix.
+
+    """
+    gating_dim = 2 if only_position else 4
+    gating_threshold = kalman_filter.chi2inv95[gating_dim]
+    measurements = np.asarray(
+        [detections[i].to_xyah() for i in detection_indices])
+    for row, track_idx in enumerate(track_indices):
+        track = tracks[track_idx]
+        gating_distance = kf.gating_distance(
+            track.mean, track.covariance, measurements, only_position)
+        cost_matrix[row, gating_distance > gating_threshold] = gated_cost
+    return cost_matrix

+# vim: expandtab:ts=4:sw=4
+import numpy as np
+
+
+def _pdist(a, b):
+    """Compute pair-wise squared distance between points in `a` and `b`.
+
+    Parameters
+    ----------
+    a : array_like
+        An NxM matrix of N samples of dimensionality M.
+    b : array_like
+        An LxM matrix of L samples of dimensionality M.
+
+    Returns
+    -------
+    ndarray
+        Returns a matrix of size len(a), len(b) such that eleement (i, j)
+        contains the squared distance between `a[i]` and `b[j]`.
+
+    """
+    a, b = np.asarray(a), np.asarray(b)
+    if len(a) == 0 or len(b) == 0:
+        return np.zeros((len(a), len(b)))
+    a2, b2 = np.square(a).sum(axis=1), np.square(b).sum(axis=1)
+    r2 = -2. * np.dot(a, b.T) + a2[:, None] + b2[None, :]
+    r2 = np.clip(r2, 0., float(np.inf))
+    return r2
+
+
+def _cosine_distance(a, b, data_is_normalized=False):
+    """Compute pair-wise cosine distance between points in `a` and `b`.
+
+    Parameters
+    ----------
+    a : array_like
+        An NxM matrix of N samples of dimensionality M.
+    b : array_like
+        An LxM matrix of L samples of dimensionality M.
+    data_is_normalized : Optional[bool]
+        If True, assumes rows in a and b are unit length vectors.
+        Otherwise, a and b are explicitly normalized to lenght 1.
+
+    Returns
+    -------
+    ndarray
+        Returns a matrix of size len(a), len(b) such that eleement (i, j)
+        contains the squared distance between `a[i]` and `b[j]`.
+
+    """
+    if not data_is_normalized:
+        a = np.asarray(a) / np.linalg.norm(a, axis=1, keepdims=True)
+        b = np.asarray(b) / np.linalg.norm(b, axis=1, keepdims=True)
+    return 1. - np.dot(a, b.T)
+
+
+def _nn_euclidean_distance(x, y):
+    """ Helper function for nearest neighbor distance metric (Euclidean).
+
+    Parameters
+    ----------
+    x : ndarray
+        A matrix of N row-vectors (sample points).
+    y : ndarray
+        A matrix of M row-vectors (query points).
+
+    Returns
+    -------
+    ndarray
+        A vector of length M that contains for each entry in `y` the
+        smallest Euclidean distance to a sample in `x`.
+
+    """
+    distances = _pdist(x, y)
+    return np.maximum(0.0, distances.min(axis=0))
+
+
+def _nn_cosine_distance(x, y):
+    """ Helper function for nearest neighbor distance metric (cosine).
+
+    Parameters
+    ----------
+    x : ndarray
+        A matrix of N row-vectors (sample points).
+    y : ndarray
+        A matrix of M row-vectors (query points).
+
+    Returns
+    -------
+    ndarray
+        A vector of length M that contains for each entry in `y` the
+        smallest cosine distance to a sample in `x`.
+
+    """
+    distances = _cosine_distance(x, y)
+    return distances.min(axis=0)
+
+
+class NearestNeighborDistanceMetric(object):
+    """
+    A nearest neighbor distance metric that, for each target, returns
+    the closest distance to any sample that has been observed so far.
+
+    Parameters
+    ----------
+    metric : str
+        Either "euclidean" or "cosine".
+    matching_threshold: float
+        The matching threshold. Samples with larger distance are considered an
+        invalid match.
+    budget : Optional[int]
+        If not None, fix samples per class to at most this number. Removes
+        the oldest samples when the budget is reached.
+
+    Attributes
+    ----------
+    samples : Dict[int -> List[ndarray]]
+        A dictionary that maps from target identities to the list of samples
+        that have been observed so far.
+
+    """
+
+    def __init__(self, metric, matching_threshold, budget=None):
+
+
+        if metric == "euclidean":
+            self._metric = _nn_euclidean_distance
+        elif metric == "cosine":
+            self._metric = _nn_cosine_distance
+        else:
+            raise ValueError(
+                "Invalid metric; must be either 'euclidean' or 'cosine'")
+        self.matching_threshold = matching_threshold
+        self.budget = budget
+        self.samples = {}
+
+    def partial_fit(self, features, targets, active_targets):
+        """Update the distance metric with new data.
+
+        Parameters
+        ----------
+        features : ndarray
+            An NxM matrix of N features of dimensionality M.
+        targets : ndarray
+            An integer array of associated target identities.
+        active_targets : List[int]
+            A list of targets that are currently present in the scene.
+
+        """
+        for feature, target in zip(features, targets):
+            self.samples.setdefault(target, []).append(feature)
+            if self.budget is not None:
+                self.samples[target] = self.samples[target][-self.budget:]
+        self.samples = {k: self.samples[k] for k in active_targets}
+
+    def distance(self, features, targets):
+        """Compute distance between features and targets.
+
+        Parameters
+        ----------
+        features : ndarray
+            An NxM matrix of N features of dimensionality M.
+        targets : List[int]
+            A list of targets to match the given `features` against.
+
+        Returns
+        -------
+        ndarray
+            Returns a cost matrix of shape len(targets), len(features), where
+            element (i, j) contains the closest squared distance between
+            `targets[i]` and `features[j]`.
+
+        """
+        cost_matrix = np.zeros((len(targets), len(features)))
+        for i, target in enumerate(targets):
+            cost_matrix[i, :] = self._metric(self.samples[target], features)
+        return cost_matrix

+# vim: expandtab:ts=4:sw=4
+import numpy as np
+import cv2
+
+
+def non_max_suppression(boxes, max_bbox_overlap, scores=None):
+    """Suppress overlapping detections.
+
+    Original code from [1]_ has been adapted to include confidence score.
+
+    .. [1] http://www.pyimagesearch.com/2015/02/16/
+           faster-non-maximum-suppression-python/
+
+    Examples
+    --------
+
+        >>> boxes = [d.roi for d in detections]
+        >>> scores = [d.confidence for d in detections]
+        >>> indices = non_max_suppression(boxes, max_bbox_overlap, scores)
+        >>> detections = [detections[i] for i in indices]
+
+    Parameters
+    ----------
+    boxes : ndarray
+        Array of ROIs (x, y, width, height).
+    max_bbox_overlap : float
+        ROIs that overlap more than this values are suppressed.
+    scores : Optional[array_like]
+        Detector confidence score.
+
+    Returns
+    -------
+    List[int]
+        Returns indices of detections that have survived non-maxima suppression.
+
+    """
+    if len(boxes) == 0:
+        return []
+
+    boxes = boxes.astype(np.float)
+    pick = []
+
+    x1 = boxes[:, 0]
+    y1 = boxes[:, 1]
+    x2 = boxes[:, 2] + boxes[:, 0]
+    y2 = boxes[:, 3] + boxes[:, 1]
+
+    area = (x2 - x1 + 1) * (y2 - y1 + 1)
+    if scores is not None:
+        idxs = np.argsort(scores)
+    else:
+        idxs = np.argsort(y2)
+
+    while len(idxs) > 0:
+        last = len(idxs) - 1
+        i = idxs[last]
+        pick.append(i)
+
+        xx1 = np.maximum(x1[i], x1[idxs[:last]])
+        yy1 = np.maximum(y1[i], y1[idxs[:last]])
+        xx2 = np.minimum(x2[i], x2[idxs[:last]])
+        yy2 = np.minimum(y2[i], y2[idxs[:last]])
+
+        w = np.maximum(0, xx2 - xx1 + 1)
+        h = np.maximum(0, yy2 - yy1 + 1)
+
+        overlap = (w * h) / area[idxs[:last]]
+
+        idxs = np.delete(
+            idxs, np.concatenate(
+                ([last], np.where(overlap > max_bbox_overlap)[0])))
+
+    return pick

+# vim: expandtab:ts=4:sw=4
+
+
+class TrackState:
+    """
+    Enumeration type for the single target track state. Newly created tracks are
+    classified as `tentative` until enough evidence has been collected. Then,
+    the track state is changed to `confirmed`. Tracks that are no longer alive
+    are classified as `deleted` to mark them for removal from the set of active
+    tracks.
+
+    """
+
+    Tentative = 1
+    Confirmed = 2
+    Deleted = 3
+
+
+class Track:
+    """
+    A single target track with state space `(x, y, a, h)` and associated
+    velocities, where `(x, y)` is the center of the bounding box, `a` is the
+    aspect ratio and `h` is the height.
+
+    Parameters
+    ----------
+    mean : ndarray
+        Mean vector of the initial state distribution.
+    covariance : ndarray
+        Covariance matrix of the initial state distribution.
+    track_id : int
+        A unique track identifier.
+    n_init : int
+        Number of consecutive detections before the track is confirmed. The
+        track state is set to `Deleted` if a miss occurs within the first
+        `n_init` frames.
+    max_age : int
+        The maximum number of consecutive misses before the track state is
+        set to `Deleted`.
+    feature : Optional[ndarray]
+        Feature vector of the detection this track originates from. If not None,
+        this feature is added to the `features` cache.
+
+    Attributes
+    ----------
+    mean : ndarray
+        Mean vector of the initial state distribution.
+    covariance : ndarray
+        Covariance matrix of the initial state distribution.
+    track_id : int
+        A unique track identifier.
+    hits : int
+        Total number of measurement updates.
+    age : int
+        Total number of frames since first occurance.
+    time_since_update : int
+        Total number of frames since last measurement update.
+    state : TrackState
+        The current track state.
+    features : List[ndarray]
+        A cache of features. On each measurement update, the associated feature
+        vector is added to this list.
+
+    """
+
+    def __init__(self, mean, covariance, track_id, n_init, max_age,
+                 feature=None):
+        self.mean = mean
+        self.covariance = covariance
+        self.track_id = track_id
+        self.hits = 1
+        self.age = 1
+        self.time_since_update = 0
+
+        self.state = TrackState.Tentative
+        self.features = []
+        if feature is not None:
+            self.features.append(feature)
+
+        self._n_init = n_init
+        self._max_age = max_age
+
+    def to_tlwh(self):
+        """Get current position in bounding box format `(top left x, top left y,
+        width, height)`.
+
+        Returns
+        -------
+        ndarray
+            The bounding box.
+
+        """
+        ret = self.mean[:4].copy()
+        ret[2] *= ret[3]
+        ret[:2] -= ret[2:] / 2
+        return ret
+
+    def to_tlbr(self):
+        """Get current position in bounding box format `(min x, miny, max x,
+        max y)`.
+
+        Returns
+        -------
+        ndarray
+            The bounding box.
+
+        """
+        ret = self.to_tlwh()
+        ret[2:] = ret[:2] + ret[2:]
+        return ret
+
+    def predict(self, kf):
+        """Propagate the state distribution to the current time step using a
+        Kalman filter prediction step.
+
+        Parameters
+        ----------
+        kf : kalman_filter.KalmanFilter
+            The Kalman filter.
+
+        """
+        self.mean, self.covariance = kf.predict(self.mean, self.covariance)
+        self.age += 1
+        self.time_since_update += 1
+
+    def update(self, kf, detection):
+        """Perform Kalman filter measurement update step and update the feature
+        cache.
+
+        Parameters
+        ----------
+        kf : kalman_filter.KalmanFilter
+            The Kalman filter.
+        detection : Detection
+            The associated detection.
+
+        """
+        self.mean, self.covariance = kf.update(
+            self.mean, self.covariance, detection.to_xyah())
+        self.features.append(detection.feature)
+
+        self.hits += 1
+        self.time_since_update = 0
+        if self.state == TrackState.Tentative and self.hits >= self._n_init:
+            self.state = TrackState.Confirmed
+
+    def mark_missed(self):
+        """Mark this track as missed (no association at the current time step).
+        """
+        if self.state == TrackState.Tentative:
+            self.state = TrackState.Deleted
+        elif self.time_since_update > self._max_age:
+            self.state = TrackState.Deleted
+
+    def is_tentative(self):
+        """Returns True if this track is tentative (unconfirmed).
+        """
+        return self.state == TrackState.Tentative
+
+    def is_confirmed(self):
+        """Returns True if this track is confirmed."""
+        return self.state == TrackState.Confirmed
+
+    def is_deleted(self):
+        """Returns True if this track is dead and should be deleted."""
+        return self.state == TrackState.Deleted

+# vim: expandtab:ts=4:sw=4
+from __future__ import absolute_import
+import numpy as np
+from . import kalman_filter
+from . import linear_assignment
+from . import iou_matching
+from .track import Track
+
+
+class Tracker:
+    """
+    This is the multi-target tracker.
+
+    Parameters
+    ----------
+    metric : nn_matching.NearestNeighborDistanceMetric
+        A distance metric for measurement-to-track association.
+    max_age : int
+        Maximum number of missed misses before a track is deleted.
+    n_init : int
+        Number of consecutive detections before the track is confirmed. The
+        track state is set to `Deleted` if a miss occurs within the first
+        `n_init` frames.
+
+    Attributes
+    ----------
+    metric : nn_matching.NearestNeighborDistanceMetric
+        The distance metric used for measurement to track association.
+    max_age : int
+        Maximum number of missed misses before a track is deleted.
+    n_init : int
+        Number of frames that a track remains in initialization phase.
+    kf : kalman_filter.KalmanFilter
+        A Kalman filter to filter target trajectories in image space.
+    tracks : List[Track]
+        The list of active tracks at the current time step.
+
+    """
+
+    def __init__(self, metric, max_iou_distance=0.7, max_age=30, n_init=3):
+        self.metric = metric
+        self.max_iou_distance = max_iou_distance
+        self.max_age = max_age
+        self.n_init = n_init
+
+        self.kf = kalman_filter.KalmanFilter()
+        self.tracks = []
+        self._next_id = 1
+
+    def predict(self):
+        """Propagate track state distributions one time step forward.
+
+        This function should be called once every time step, before `update`.
+        """
+        for track in self.tracks:
+            track.predict(self.kf)
+
+    def update(self, detections):
+        """Perform measurement update and track management.
+
+        Parameters
+        ----------
+        detections : List[deep_sort.detection.Detection]
+            A list of detections at the current time step.
+
+        """
+        # Run matching cascade.
+        matches, unmatched_tracks, unmatched_detections = \
+            self._match(detections)
+
+        # Update track set.
+        for track_idx, detection_idx in matches:
+            self.tracks[track_idx].update(
+                self.kf, detections[detection_idx])
+        for track_idx in unmatched_tracks:
+            self.tracks[track_idx].mark_missed()
+        for detection_idx in unmatched_detections:
+            self._initiate_track(detections[detection_idx])
+        self.tracks = [t for t in self.tracks if not t.is_deleted()]
+
+        # Update distance metric.
+        active_targets = [t.track_id for t in self.tracks if t.is_confirmed()]
+        features, targets = [], []
+        for track in self.tracks:
+            if not track.is_confirmed():
+                continue
+            features += track.features
+            targets += [track.track_id for _ in track.features]
+            track.features = []
+        self.metric.partial_fit(
+            np.asarray(features), np.asarray(targets), active_targets)
+
+    def _match(self, detections):
+
+        def gated_metric(tracks, dets, track_indices, detection_indices):
+            features = np.array([dets[i].feature for i in detection_indices])
+            targets = np.array([tracks[i].track_id for i in track_indices])
+            cost_matrix = self.metric.distance(features, targets)
+            cost_matrix = linear_assignment.gate_cost_matrix(
+                self.kf, cost_matrix, tracks, dets, track_indices,
+                detection_indices)
+
+            return cost_matrix
+
+        # Split track set into confirmed and unconfirmed tracks.
+        confirmed_tracks = [
+            i for i, t in enumerate(self.tracks) if t.is_confirmed()]
+        unconfirmed_tracks = [
+            i for i, t in enumerate(self.tracks) if not t.is_confirmed()]
+
+        # Associate confirmed tracks using appearance features.
+        matches_a, unmatched_tracks_a, unmatched_detections = \
+            linear_assignment.matching_cascade(
+                gated_metric, self.metric.matching_threshold, self.max_age,
+                self.tracks, detections, confirmed_tracks)
+
+        # Associate remaining tracks together with unconfirmed tracks using IOU.
+        iou_track_candidates = unconfirmed_tracks + [
+            k for k in unmatched_tracks_a if
+            self.tracks[k].time_since_update == 1]
+        unmatched_tracks_a = [
+            k for k in unmatched_tracks_a if
+            self.tracks[k].time_since_update != 1]
+        matches_b, unmatched_tracks_b, unmatched_detections = \
+            linear_assignment.min_cost_matching(
+                iou_matching.iou_cost, self.max_iou_distance, self.tracks,
+                detections, iou_track_candidates, unmatched_detections)
+
+        matches = matches_a + matches_b
+        unmatched_tracks = list(set(unmatched_tracks_a + unmatched_tracks_b))
+        return matches, unmatched_tracks, unmatched_detections
+
+    def _initiate_track(self, detection):
+        mean, covariance = self.kf.initiate(detection.to_xyah())
+        self.tracks.append(Track(
+            mean, covariance, self._next_id, self.n_init, self.max_age,
+            detection.feature))
+        self._next_id += 1

+#! /usr/bin/env python
+# -*- coding: utf-8 -*-
+
+from __future__ import division, print_function, absolute_import
+import sys
+#sys.path.remove('/opt/ros/kinetic/lib/python2.7/dist-packages')
+import os
+import datetime
+from timeit import time
+import warnings
+import cv2
+import numpy as np
+import argparse
+from sklearn.cluster import KMeans
+import matplotlib.pyplot as plt
+from PIL import Image
+from yolo import YOLO
+import matplotlib.cbook as cbook
+
+from deep_sort import preprocessing
+from deep_sort import nn_matching
+from deep_sort.detection import Detection
+from deep_sort.tracker import Tracker
+from tools import generate_detections as gdet
+from deep_sort.detection import Detection as ddet
+from collections import deque
+from keras import backend
+import tensorflow as tf
+from tensorflow.compat.v1 import InteractiveSession
+import pandas as pd
+import json
+
+def createKmeans(file_name):
+    data = json.load(open(os.getcwd() + '/../deep_sort_yolov4/output/'+ file_name +'_xy.json'))
+
+    data = pd.DataFrame(data["data"])
+
+    xy = data[['x','y']]
+
+    km = KMeans(n_clusters=5)
+    km.fit(xy)
+
+    predict = pd.DataFrame(km.predict(xy))
+    predict.columns=['predict']
+    r = pd.concat([xy,predict],axis=1)
+
+    plt.scatter(r['x'],r['y'],c=r['predict'],alpha=0.3, s=200)
+
+
+    imageFile = cbook.get_sample_data(os.getcwd() + '/../deep_sort_yolov4/output/'+ file_name + '_img.png')
+    image = plt.imread(imageFile)
+    plt.imshow(image)
+    plt.savefig(os.getcwd() + '/public/data/'+file_name+'_kmeans.png')
+
+def draw_border(img, pt1, pt2, color, thickness, r, d):
+    x1,y1 = pt1
+    x2,y2 = pt2
+ 
+    # Top left
+    cv2.line(img, (x1 + r, y1), (x1 + r + d, y1), color, thickness)
+    cv2.line(img, (x1, y1 + r), (x1, y1 + r + d), color, thickness)
+    cv2.ellipse(img, (x1 + r, y1 + r), (r, r), 180, 0, 90, color, thickness)
+ 
+    # Top right
+    cv2.line(img, (x2 - r, y1), (x2 - r - d, y1), color, thickness)
+    cv2.line(img, (x2, y1 + r), (x2, y1 + r + d), color, thickness)
+    cv2.ellipse(img, (x2 - r, y1 + r), (r, r), 270, 0, 90, color, thickness)
+ 
+    # Bottom left
+    cv2.line(img, (x1 + r, y2), (x1 + r + d, y2), color, thickness)
+    cv2.line(img, (x1, y2 - r), (x1, y2 - r - d), color, thickness)
+    cv2.ellipse(img, (x1 + r, y2 - r), (r, r), 90, 0, 90, color, thickness)
+ 
+    # Bottom right
+    cv2.line(img, (x2 - r, y2), (x2 - r - d, y2), color, thickness)
+    cv2.line(img, (x2, y2 - r), (x2, y2 - r - d), color, thickness)
+    cv2.ellipse(img, (x2 - r, y2 - r), (r, r), 0, 0, 90, color, thickness)
+
+config = tf.ConfigProto()
+config.gpu_options.allow_growth = True
+session = InteractiveSession(config=config)
+
+
+#ap = argparse.ArgumentParser()
+#ap.add_argument("-i", "--input",help="path to input video", default = "./test_video/test3.mp4")
+#ap.add_argument("-c", "--class",help="name of class", default = "person")
+#args = vars(ap.parse_args())
+
+pts = [deque(maxlen=30) for _ in range(9999)]
+warnings.filterwarnings('ignore')
+
+# initialize a list of colors to represent each possible class label
+np.random.seed(100)
+COLORS = np.random.randint(0, 255, size=(200, 3),
+	dtype="uint8")
+#list = [[] for _ in range(100)]
+
+def main(yolo):
+    df1 = pd.DataFrame(columns=['x','y','id','time','s'])
+    df2 = pd.DataFrame(columns=['total','now','time','s'])
+    start = datetime.datetime.now()
+    max_cosine_distance = 0.3
+    nn_budget = None
+    nms_max_overlap = 1.0
+
+    counter = []
+    #deep_sort
+    model_filename = os.getcwd() + '/../deep_sort_yolov4/model_data/market1501.pb'
+    encoder = gdet.create_box_encoder(model_filename,batch_size=1)
+
+    find_objects = ['person']
+    metric = nn_matching.NearestNeighborDistanceMetric("cosine", max_cosine_distance, nn_budget)
+    tracker = Tracker(metric)
+
+    writeVideo_flag = True
+    video_capture = cv2.VideoCapture(os.getcwd() + '/../deep_sort_yolov4/' + sys.argv[1])
+
+    file_name = sys.argv[1].split('.')[0]
+    if writeVideo_flag:
+    # Define the codec and create VideoWriter object
+        w = int(video_capture.get(3))
+        h = int(video_capture.get(4))
+        total_frame = int(video_capture.get(cv2.CAP_PROP_FRAME_COUNT))
+        fourcc = cv2.VideoWriter_fourcc(*'MP4V')
+        out = cv2.VideoWriter(os.getcwd() + '/public/data/'+file_name+'.mp4', fourcc, 15, (w, h))
+        list_file = open(os.getcwd() + '/../deep_sort_yolov4/detection_rslt.txt', 'w')
+        frame_index = -1
+
+    fps = 0.0
+    test = 1
+    while True:
+
+        ret, frame = video_capture.read()  # frame shape 640*480*3
+        if ret != True:
+            break
+        cFrame = int(video_capture.get(cv2.CAP_PROP_POS_MSEC)/76)
+        #temp = pd.DataFrame({'num':cFrame},index = [0])
+        #temp.to_csv(os.getcwd() + '/../deep_sort_yolov4/output/temp.csv')
+        t1 = time.time()
+        if(test == 1 ):
+            cv2.imwrite(os.getcwd() + '/../deep_sort_yolov4/output/' + file_name + '_img.png',frame)
+            test +=1
+
+        #image = Image.fromarray(frame)
+        image = Image.fromarray(frame[...,::-1]) #bgr to rgb
+        boxs, confidence, class_names = yolo.detect_image(image)
+        features = encoder(frame,boxs)
+        # score to 1.0 here).
+        detections = [Detection(bbox, 1.0, feature) for bbox, feature in zip(boxs, features)]
+        # Run non-maxima suppression.
+        boxes = np.array([d.tlwh for d in detections])
+        scores = np.array([d.confidence for d in detections])
+        indices = preprocessing.non_max_suppression(boxes, nms_max_overlap, scores)
+        detections = [detections[i] for i in indices]
+        cs = float(59)*int(cFrame)/total_frame
+        t = start + datetime.timedelta(milliseconds=cs*1000)
+        time_format = "%Y-%m-%d %H:%M:%S.%f"
+        time_str = t.strftime(time_format)
+        cTime = time_str
+
+        # Call the tracker
+        tracker.predict()
+        tracker.update(detections)
+
+        i = int(0)
+        indexIDs = []
+        c = []
+        boxes = []
+
+        for track in tracker.tracks:
+            if not track.is_confirmed() or track.time_since_update > 1:
+                continue
+            #boxes.append([track[0], track[1], track[2], track[3]])
+            indexIDs.append(int(track.track_id))
+            counter.append(int(track.track_id))
+            bbox = track.to_tlbr()
+            color = [int(c) for c in COLORS[indexIDs[i] % len(COLORS)]]
+            #print(frame_index)
+            list_file.write(str(frame_index)+',')
+            list_file.write(str(track.track_id)+',')
+            center = (int(((bbox[0])+(bbox[2]))/2),int(((bbox[1])+(bbox[3]))/2))
+            r = (int(bbox[2]) - int(bbox[0]))/2
+            draw_border(frame,(int(((bbox[0])+(bbox[2]))/2) - int(r), int(((bbox[1])+(bbox[3]))/2) - int(r)), (int(((bbox[0])+(bbox[2]))/2) + int(r), int(((bbox[1])+(bbox[3]))/2) + int(r)),(color),4,int(r/5),int(r/5))
+            #cv2.rectangle(frame, (int(bbox[0]), int(bbox[1])), (int(bbox[2]), int(bbox[3])),(color), 3)
+            b0 = str(bbox[0])#.split('.')[0] + '.' + str(bbox[0]).split('.')[0][:1]
+            b1 = str(bbox[1])#.split('.')[0] + '.' + str(bbox[1]).split('.')[0][:1]
+            b2 = str(bbox[2]-bbox[0])#.split('.')[0] + '.' + str(bbox[3]).split('.')[0][:1]
+            b3 = str(bbox[3]-bbox[1])
+
+            list_file.write(str(b0) + ','+str(b1) + ','+str(b2) + ','+str(b3))
+            #print(str(track.track_id))
+            list_file.write('\n')
+            #list_file.write(str(track.track_id)+',')
+            #cv2.putText(frame,str(track.track_id),(int(bbox[0]), int(bbox[1] + 30)),0, 5e-3 * 150, (color),2)
+            cv2.putText(frame,str(track.track_id),(int(((bbox[0])+(bbox[2]))/2), int(((bbox[1])+(bbox[3]))/2 + 30)),0, 5e-3 * 150, (color),2)
+
+            if len(class_names) > 0:
+               class_name = class_names[0]
+               cv2.putText(frame, str(class_names[0]),(int(((bbox[0])+(bbox[2]))/2), int(((bbox[1])+(bbox[3]))/2 + 10)),0, 5e-3 * 150, (color),2)
+
+            i += 1
+            #bbox_center_point(x,y)
+            #track_id[center]
+            x,y = int(((bbox[0])+(bbox[2]))/2),int(((bbox[1])+(bbox[3]))/2)
+            # x, y
+            if (cFrame%10==0):
+                df1 = df1.append({'x':x,'y':y,'id':int(track.track_id),'time':cTime,'s': cs},ignore_index=True)
+
+            pts[track.track_id].append(center)
+
+            thickness = 5
+            #center point
+            #cv2.circle(frame,  (center), 1, color, thickness)
+
+			# draw motion path
+            # for j in range(1, len(pts[track.track_id])):
+            #     if pts[track.track_id][j - 1] is None or pts[track.track_id][j] is None:
+            #        continue
+            #     thickness = int(np.sqrt(64 / float(j + 1)) * 2)
+                #cv2.line(frame,(pts[track.track_id][j-1]), (pts[track.track_id][j]),(color),thickness)
+                #cv2.putText(frame, str(class_names[j]),(int(bbox[0]), int(bbox[1] -20)),0, 5e-3 * 150, (255,255,255),2)
+
+        count = len(set(counter))
+        if (cFrame%10==0):
+            df2 = df2.append({'total':count,'now':i,'time':cTime,'s': cs},ignore_index=True)
+        cv2.putText(frame, "Total Pedestrian Counter: "+str(count),(int(30), int(105)),0, 5e-3 * 150, (0,255,0),2)
+        cv2.putText(frame, "Current Pedestrian Counter: "+str(i),(int(30), int(80)),0, 5e-3 * 150, (0,255,0),2)
+        #cv2.putText(frame, "FPS: %f"%(fps),(int(20), int(40)),0, 5e-3 * 200, (0,255,0),3)
+        cv2.putText(frame, "Time: " + str(cTime),(int(30), int(55)),0, 5e-3 * 150, (0,255,0),3)
+        cv2.namedWindow("YOLO4_Deep_SORT", 0);
+        cv2.resizeWindow('YOLO4_Deep_SORT', 1024, 768);
+        cv2.imshow('YOLO4_Deep_SORT', frame)
+
+
+        if writeVideo_flag:
+            # save a frame
+            out.write(frame)
+            frame_index = frame_index + 1
+
+
+        fps  = ( fps + (1./(time.time()-t1)) ) / 2
+        out.write(frame)
+        frame_index = frame_index + 1
+
+        # Press Q to stop!
+        if cv2.waitKey(1) & 0xFF == ord('q'):
+            break
+    df1.to_json(os.getcwd() + '/../deep_sort_yolov4/output/' + file_name + '_xy.json',orient='table')
+    df2.to_json(os.getcwd() + '/../deep_sort_yolov4/output/' + file_name + '_count.json',orient='table')
+    createKmeans(file_name)
+    print(" ")
+    print("[Finish]")
+    end = time.time()
+
+	#print("[INFO]: model_image_size = (960, 960)")
+    video_capture.release()
+    if writeVideo_flag:
+        out.release()
+        list_file.close()
+    cv2.destroyAllWindows()
+
+if __name__ == '__main__':
+    main(YOLO())

+person
+bicycle
+car
+motorbike
+aeroplane
+bus
+train
+truck
+boat
+traffic light
+fire hydrant
+stop sign
+parking meter
+bench
+bird
+cat
+dog
+horse
+sheep
+cow
+elephant
+bear
+zebra
+giraffe
+backpack
+umbrella
+handbag
+tie
+suitcase
+frisbee
+skis
+snowboard
+sports ball
+kite
+baseball bat
+baseball glove
+skateboard
+surfboard
+tennis racket
+bottle
+wine glass
+cup
+fork
+knife
+spoon
+bowl
+banana
+apple
+sandwich
+orange
+broccoli
+carrot
+hot dog
+pizza
+donut
+cake
+chair
+sofa
+pottedplant
+bed
+diningtable
+toilet
+tvmonitor
+laptop
+mouse
+remote
+keyboard
+cell phone
+microwave
+oven
+toaster
+sink
+refrigerator
+book
+clock
+vase
+scissors
+teddy bear
+hair drier
+toothbrush

+person

+12, 16,  19, 36,  40, 28,  36, 75,  76, 55,  72, 146,  142, 110,  192, 243,  459, 401

+Output_video
+

+{"schema":{"fields":[{"name":"index","type":"integer"},{"name":"total","type":"string"},{"name":"now","type":"string"},{"name":"time","type":"string"},{"name":"s","type":"number"}],"primaryKey":["index"],"pandas_version":"0.20.0"},"data":[{"index":0,"total":0,"now":0,"time":"2020-06-17 03:27:10.288851","s":0.0},{"index":1,"total":11,"now":7,"time":"2020-06-17 03:27:11.051125","s":0.7622739018},{"index":2,"total":13,"now":8,"time":"2020-06-17 03:27:11.813399","s":1.5245478036},{"index":3,"total":15,"now":10,"time":"2020-06-17 03:27:12.575673","s":2.2868217054},{"index":4,"total":21,"now":13,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":5,"total":22,"now":13,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":6,"total":23,"now":12,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":7,"total":24,"now":11,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":8,"total":24,"now":7,"time":"2020-06-17 03:27:16.387042","s":6.0981912145},{"index":9,"total":25,"now":8,"time":"2020-06-17 03:27:17.149316","s":6.8604651163},{"index":10,"total":26,"now":9,"time":"2020-06-17 03:27:17.911590","s":7.6227390181},{"index":11,"total":26,"now":10,"time":"2020-06-17 03:27:18.673864","s":8.3850129199},{"index":12,"total":26,"now":7,"time":"2020-06-17 03:27:19.436138","s":9.1472868217},{"index":13,"total":29,"now":8,"time":"2020-06-17 03:27:20.198412","s":9.9095607235},{"index":14,"total":31,"now":11,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":15,"total":31,"now":10,"time":"2020-06-17 03:27:21.722960","s":11.4341085271},{"index":16,"total":32,"now":13,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":17,"total":33,"now":11,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":18,"total":33,"now":11,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":19,"total":35,"now":11,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":20,"total":37,"now":9,"time":"2020-06-17 03:27:25.534329","s":15.2454780362},{"index":21,"total":38,"now":10,"time":"2020-06-17 03:27:26.296603","s":16.007751938},{"index":22,"total":39,"now":11,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":23,"total":41,"now":12,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":24,"total":42,"now":11,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":25,"total":42,"now":12,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":26,"total":42,"now":10,"time":"2020-06-17 03:27:30.107972","s":19.819121447},{"index":27,"total":42,"now":8,"time":"2020-06-17 03:27:30.870246","s":20.5813953488},{"index":28,"total":43,"now":9,"time":"2020-06-17 03:27:31.632520","s":21.3436692506},{"index":29,"total":45,"now":11,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":30,"total":45,"now":9,"time":"2020-06-17 03:27:33.157068","s":22.8682170543},{"index":31,"total":46,"now":11,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":32,"total":46,"now":11,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":33,"total":46,"now":11,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":34,"total":47,"now":10,"time":"2020-06-17 03:27:36.206164","s":25.9173126615},{"index":35,"total":47,"now":9,"time":"2020-06-17 03:27:36.968438","s":26.6795865633},{"index":36,"total":48,"now":10,"time":"2020-06-17 03:27:37.730711","s":27.4418604651},{"index":37,"total":48,"now":10,"time":"2020-06-17 03:27:38.492985","s":28.2041343669},{"index":38,"total":50,"now":10,"time":"2020-06-17 03:27:39.255259","s":28.9664082687},{"index":39,"total":51,"now":10,"time":"2020-06-17 03:27:40.017533","s":29.7286821705},{"index":40,"total":52,"now":9,"time":"2020-06-17 03:27:40.779807","s":30.4909560724},{"index":41,"total":52,"now":8,"time":"2020-06-17 03:27:41.542081","s":31.2532299742},{"index":42,"total":53,"now":7,"time":"2020-06-17 03:27:42.304355","s":32.015503876},{"index":43,"total":55,"now":8,"time":"2020-06-17 03:27:43.066629","s":32.7777777778},{"index":44,"total":56,"now":8,"time":"2020-06-17 03:27:43.828903","s":33.5400516796},{"index":45,"total":56,"now":10,"time":"2020-06-17 03:27:44.591177","s":34.3023255814},{"index":46,"total":56,"now":9,"time":"2020-06-17 03:27:45.353450","s":35.0645994832},{"index":47,"total":57,"now":10,"time":"2020-06-17 03:27:46.115724","s":35.826873385},{"index":48,"total":57,"now":8,"time":"2020-06-17 03:27:46.877998","s":36.5891472868},{"index":49,"total":58,"now":10,"time":"2020-06-17 03:27:47.640272","s":37.3514211886},{"index":50,"total":58,"now":8,"time":"2020-06-17 03:27:48.402546","s":38.1136950904},{"index":51,"total":60,"now":8,"time":"2020-06-17 03:27:49.164820","s":38.8759689922},{"index":52,"total":63,"now":12,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":53,"total":63,"now":12,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":54,"total":63,"now":8,"time":"2020-06-17 03:27:51.451642","s":41.1627906977},{"index":55,"total":63,"now":10,"time":"2020-06-17 03:27:52.213916","s":41.9250645995},{"index":56,"total":64,"now":11,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":57,"total":64,"now":9,"time":"2020-06-17 03:27:53.738463","s":43.4496124031},{"index":58,"total":64,"now":10,"time":"2020-06-17 03:27:54.500737","s":44.2118863049},{"index":59,"total":64,"now":9,"time":"2020-06-17 03:27:55.263011","s":44.9741602067},{"index":60,"total":64,"now":8,"time":"2020-06-17 03:27:56.025285","s":45.7364341085},{"index":61,"total":64,"now":9,"time":"2020-06-17 03:27:56.787559","s":46.4987080103},{"index":62,"total":65,"now":8,"time":"2020-06-17 03:27:57.549833","s":47.2609819121},{"index":63,"total":66,"now":9,"time":"2020-06-17 03:27:58.312107","s":48.023255814},{"index":64,"total":67,"now":10,"time":"2020-06-17 03:27:59.074381","s":48.7855297158},{"index":65,"total":68,"now":10,"time":"2020-06-17 03:27:59.836655","s":49.5478036176},{"index":66,"total":68,"now":9,"time":"2020-06-17 03:28:00.598929","s":50.3100775194},{"index":67,"total":69,"now":9,"time":"2020-06-17 03:28:01.361202","s":51.0723514212},{"index":68,"total":69,"now":8,"time":"2020-06-17 03:28:02.123476","s":51.834625323},{"index":69,"total":69,"now":10,"time":"2020-06-17 03:28:02.885750","s":52.5968992248},{"index":70,"total":70,"now":8,"time":"2020-06-17 03:28:03.648024","s":53.3591731266},{"index":71,"total":71,"now":9,"time":"2020-06-17 03:28:04.410298","s":54.1214470284},{"index":72,"total":73,"now":11,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":73,"total":73,"now":10,"time":"2020-06-17 03:28:05.934846","s":55.645994832},{"index":74,"total":74,"now":9,"time":"2020-06-17 03:28:06.697120","s":56.4082687339},{"index":75,"total":75,"now":10,"time":"2020-06-17 03:28:07.459394","s":57.1705426357},{"index":76,"total":75,"now":7,"time":"2020-06-17 03:28:08.221668","s":57.9328165375},{"index":77,"total":77,"now":10,"time":"2020-06-17 03:28:08.983941","s":58.6950904393}]}
\ No newline at end of file

+{"schema":{"fields":[{"name":"index","type":"integer"},{"name":"x","type":"string"},{"name":"y","type":"string"},{"name":"id","type":"string"},{"name":"time","type":"string"},{"name":"s","type":"number"}],"primaryKey":["index"],"pandas_version":"0.20.0"},"data":[{"index":0,"x":902,"y":316,"id":1,"time":"2020-06-17 03:27:11.051125","s":0.7622739018},{"index":1,"x":747,"y":541,"id":2,"time":"2020-06-17 03:27:11.051125","s":0.7622739018},{"index":2,"x":75,"y":354,"id":3,"time":"2020-06-17 03:27:11.051125","s":0.7622739018},{"index":3,"x":918,"y":559,"id":4,"time":"2020-06-17 03:27:11.051125","s":0.7622739018},{"index":4,"x":652,"y":104,"id":5,"time":"2020-06-17 03:27:11.051125","s":0.7622739018},{"index":5,"x":288,"y":205,"id":7,"time":"2020-06-17 03:27:11.051125","s":0.7622739018},{"index":6,"x":290,"y":110,"id":11,"time":"2020-06-17 03:27:11.051125","s":0.7622739018},{"index":7,"x":919,"y":296,"id":1,"time":"2020-06-17 03:27:11.813399","s":1.5245478036},{"index":8,"x":766,"y":470,"id":2,"time":"2020-06-17 03:27:11.813399","s":1.5245478036},{"index":9,"x":75,"y":352,"id":3,"time":"2020-06-17 03:27:11.813399","s":1.5245478036},{"index":10,"x":927,"y":551,"id":4,"time":"2020-06-17 03:27:11.813399","s":1.5245478036},{"index":11,"x":655,"y":104,"id":5,"time":"2020-06-17 03:27:11.813399","s":1.5245478036},{"index":12,"x":285,"y":202,"id":7,"time":"2020-06-17 03:27:11.813399","s":1.5245478036},{"index":13,"x":364,"y":188,"id":8,"time":"2020-06-17 03:27:11.813399","s":1.5245478036},{"index":14,"x":112,"y":250,"id":14,"time":"2020-06-17 03:27:11.813399","s":1.5245478036},{"index":15,"x":903,"y":281,"id":1,"time":"2020-06-17 03:27:12.575673","s":2.2868217054},{"index":16,"x":827,"y":435,"id":2,"time":"2020-06-17 03:27:12.575673","s":2.2868217054},{"index":17,"x":73,"y":351,"id":3,"time":"2020-06-17 03:27:12.575673","s":2.2868217054},{"index":18,"x":926,"y":569,"id":4,"time":"2020-06-17 03:27:12.575673","s":2.2868217054},{"index":19,"x":653,"y":104,"id":5,"time":"2020-06-17 03:27:12.575673","s":2.2868217054},{"index":20,"x":273,"y":188,"id":7,"time":"2020-06-17 03:27:12.575673","s":2.2868217054},{"index":21,"x":438,"y":205,"id":8,"time":"2020-06-17 03:27:12.575673","s":2.2868217054},{"index":22,"x":139,"y":248,"id":14,"time":"2020-06-17 03:27:12.575673","s":2.2868217054},{"index":23,"x":307,"y":143,"id":20,"time":"2020-06-17 03:27:12.575673","s":2.2868217054},{"index":24,"x":1051,"y":567,"id":24,"time":"2020-06-17 03:27:12.575673","s":2.2868217054},{"index":25,"x":901,"y":227,"id":1,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":26,"x":81,"y":349,"id":3,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":27,"x":867,"y":575,"id":4,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":28,"x":650,"y":104,"id":5,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":29,"x":275,"y":193,"id":7,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":30,"x":446,"y":206,"id":8,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":31,"x":141,"y":248,"id":14,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":32,"x":293,"y":116,"id":20,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":33,"x":390,"y":180,"id":26,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":34,"x":836,"y":406,"id":27,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":35,"x":394,"y":53,"id":29,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":36,"x":1014,"y":582,"id":30,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":37,"x":947,"y":337,"id":32,"time":"2020-06-17 03:27:13.337947","s":3.0490956072},{"index":38,"x":900,"y":274,"id":1,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":39,"x":1013,"y":375,"id":2,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":40,"x":88,"y":341,"id":3,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":41,"x":794,"y":560,"id":4,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":42,"x":654,"y":105,"id":5,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":43,"x":295,"y":236,"id":7,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":44,"x":447,"y":205,"id":8,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":45,"x":111,"y":202,"id":14,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":46,"x":271,"y":97,"id":20,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":47,"x":387,"y":149,"id":26,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":48,"x":917,"y":385,"id":27,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":49,"x":1006,"y":569,"id":30,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":50,"x":340,"y":117,"id":34,"time":"2020-06-17 03:27:14.100221","s":3.811369509},{"index":51,"x":902,"y":294,"id":1,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":52,"x":92,"y":298,"id":3,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":53,"x":746,"y":566,"id":4,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":54,"x":653,"y":107,"id":5,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":55,"x":296,"y":232,"id":7,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":56,"x":394,"y":198,"id":8,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":57,"x":105,"y":204,"id":14,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":58,"x":271,"y":105,"id":20,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":59,"x":972,"y":351,"id":27,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":60,"x":1015,"y":506,"id":30,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":61,"x":336,"y":116,"id":34,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":62,"x":223,"y":186,"id":36,"time":"2020-06-17 03:27:14.862494","s":4.5736434109},{"index":63,"x":904,"y":296,"id":1,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":64,"x":96,"y":322,"id":3,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":65,"x":741,"y":567,"id":4,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":66,"x":651,"y":105,"id":5,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":67,"x":300,"y":231,"id":7,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":68,"x":387,"y":173,"id":8,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":69,"x":108,"y":202,"id":14,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":70,"x":268,"y":89,"id":20,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":71,"x":1052,"y":446,"id":30,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":72,"x":225,"y":187,"id":36,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":73,"x":476,"y":70,"id":41,"time":"2020-06-17 03:27:15.624768","s":5.3359173127},{"index":74,"x":904,"y":296,"id":1,"time":"2020-06-17 03:27:16.387042","s":6.0981912145},{"index":75,"x":112,"y":319,"id":3,"time":"2020-06-17 03:27:16.387042","s":6.0981912145},{"index":76,"x":748,"y":568,"id":4,"time":"2020-06-17 03:27:16.387042","s":6.0981912145},{"index":77,"x":632,"y":100,"id":5,"time":"2020-06-17 03:27:16.387042","s":6.0981912145},{"index":78,"x":263,"y":237,"id":7,"time":"2020-06-17 03:27:16.387042","s":6.0981912145},{"index":79,"x":1079,"y":420,"id":30,"time":"2020-06-17 03:27:16.387042","s":6.0981912145},{"index":80,"x":276,"y":97,"id":34,"time":"2020-06-17 03:27:16.387042","s":6.0981912145},{"index":81,"x":897,"y":284,"id":1,"time":"2020-06-17 03:27:17.149316","s":6.8604651163},{"index":82,"x":118,"y":319,"id":3,"time":"2020-06-17 03:27:17.149316","s":6.8604651163},{"index":83,"x":723,"y":572,"id":4,"time":"2020-06-17 03:27:17.149316","s":6.8604651163},{"index":84,"x":637,"y":102,"id":5,"time":"2020-06-17 03:27:17.149316","s":6.8604651163},{"index":85,"x":251,"y":236,"id":7,"time":"2020-06-17 03:27:17.149316","s":6.8604651163},{"index":86,"x":325,"y":148,"id":8,"time":"2020-06-17 03:27:17.149316","s":6.8604651163},{"index":87,"x":1087,"y":416,"id":30,"time":"2020-06-17 03:27:17.149316","s":6.8604651163},{"index":88,"x":276,"y":110,"id":34,"time":"2020-06-17 03:27:17.149316","s":6.8604651163},{"index":89,"x":915,"y":286,"id":1,"time":"2020-06-17 03:27:17.911590","s":7.6227390181},{"index":90,"x":120,"y":318,"id":3,"time":"2020-06-17 03:27:17.911590","s":7.6227390181},{"index":91,"x":740,"y":572,"id":4,"time":"2020-06-17 03:27:17.911590","s":7.6227390181},{"index":92,"x":614,"y":101,"id":5,"time":"2020-06-17 03:27:17.911590","s":7.6227390181},{"index":93,"x":255,"y":202,"id":7,"time":"2020-06-17 03:27:17.911590","s":7.6227390181},{"index":94,"x":314,"y":143,"id":8,"time":"2020-06-17 03:27:17.911590","s":7.6227390181},{"index":95,"x":1058,"y":406,"id":30,"time":"2020-06-17 03:27:17.911590","s":7.6227390181},{"index":96,"x":383,"y":104,"id":46,"time":"2020-06-17 03:27:17.911590","s":7.6227390181},{"index":97,"x":326,"y":112,"id":49,"time":"2020-06-17 03:27:17.911590","s":7.6227390181},{"index":98,"x":904,"y":274,"id":1,"time":"2020-06-17 03:27:18.673864","s":8.3850129199},{"index":99,"x":120,"y":318,"id":3,"time":"2020-06-17 03:27:18.673864","s":8.3850129199},{"index":100,"x":725,"y":576,"id":4,"time":"2020-06-17 03:27:18.673864","s":8.3850129199},{"index":101,"x":640,"y":109,"id":5,"time":"2020-06-17 03:27:18.673864","s":8.3850129199},{"index":102,"x":259,"y":204,"id":7,"time":"2020-06-17 03:27:18.673864","s":8.3850129199},{"index":103,"x":760,"y":183,"id":27,"time":"2020-06-17 03:27:18.673864","s":8.3850129199},{"index":104,"x":1042,"y":371,"id":30,"time":"2020-06-17 03:27:18.673864","s":8.3850129199},{"index":105,"x":284,"y":119,"id":34,"time":"2020-06-17 03:27:18.673864","s":8.3850129199},{"index":106,"x":382,"y":105,"id":46,"time":"2020-06-17 03:27:18.673864","s":8.3850129199},{"index":107,"x":324,"y":119,"id":49,"time":"2020-06-17 03:27:18.673864","s":8.3850129199},{"index":108,"x":872,"y":247,"id":1,"time":"2020-06-17 03:27:19.436138","s":9.1472868217},{"index":109,"x":120,"y":317,"id":3,"time":"2020-06-17 03:27:19.436138","s":9.1472868217},{"index":110,"x":726,"y":574,"id":4,"time":"2020-06-17 03:27:19.436138","s":9.1472868217},{"index":111,"x":693,"y":114,"id":5,"time":"2020-06-17 03:27:19.436138","s":9.1472868217},{"index":112,"x":278,"y":184,"id":7,"time":"2020-06-17 03:27:19.436138","s":9.1472868217},{"index":113,"x":1006,"y":339,"id":30,"time":"2020-06-17 03:27:19.436138","s":9.1472868217},{"index":114,"x":381,"y":107,"id":46,"time":"2020-06-17 03:27:19.436138","s":9.1472868217},{"index":115,"x":845,"y":190,"id":1,"time":"2020-06-17 03:27:20.198412","s":9.9095607235},{"index":116,"x":116,"y":314,"id":3,"time":"2020-06-17 03:27:20.198412","s":9.9095607235},{"index":117,"x":719,"y":570,"id":4,"time":"2020-06-17 03:27:20.198412","s":9.9095607235},{"index":118,"x":750,"y":132,"id":5,"time":"2020-06-17 03:27:20.198412","s":9.9095607235},{"index":119,"x":958,"y":310,"id":30,"time":"2020-06-17 03:27:20.198412","s":9.9095607235},{"index":120,"x":272,"y":96,"id":34,"time":"2020-06-17 03:27:20.198412","s":9.9095607235},{"index":121,"x":381,"y":116,"id":46,"time":"2020-06-17 03:27:20.198412","s":9.9095607235},{"index":122,"x":675,"y":120,"id":60,"time":"2020-06-17 03:27:20.198412","s":9.9095607235},{"index":123,"x":850,"y":181,"id":1,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":124,"x":116,"y":326,"id":3,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":125,"x":715,"y":569,"id":4,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":126,"x":764,"y":163,"id":5,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":127,"x":277,"y":184,"id":7,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":128,"x":955,"y":304,"id":30,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":129,"x":275,"y":91,"id":34,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":130,"x":381,"y":117,"id":46,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":131,"x":328,"y":122,"id":49,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":132,"x":586,"y":62,"id":57,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":133,"x":280,"y":228,"id":58,"time":"2020-06-17 03:27:20.960686","s":10.6718346253},{"index":134,"x":828,"y":169,"id":1,"time":"2020-06-17 03:27:21.722960","s":11.4341085271},{"index":135,"x":117,"y":317,"id":3,"time":"2020-06-17 03:27:21.722960","s":11.4341085271},{"index":136,"x":720,"y":569,"id":4,"time":"2020-06-17 03:27:21.722960","s":11.4341085271},{"index":137,"x":857,"y":236,"id":5,"time":"2020-06-17 03:27:21.722960","s":11.4341085271},{"index":138,"x":958,"y":307,"id":30,"time":"2020-06-17 03:27:21.722960","s":11.4341085271},{"index":139,"x":275,"y":95,"id":34,"time":"2020-06-17 03:27:21.722960","s":11.4341085271},{"index":140,"x":379,"y":117,"id":46,"time":"2020-06-17 03:27:21.722960","s":11.4341085271},{"index":141,"x":328,"y":115,"id":49,"time":"2020-06-17 03:27:21.722960","s":11.4341085271},{"index":142,"x":650,"y":84,"id":57,"time":"2020-06-17 03:27:21.722960","s":11.4341085271},{"index":143,"x":280,"y":195,"id":58,"time":"2020-06-17 03:27:21.722960","s":11.4341085271},{"index":144,"x":823,"y":171,"id":1,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":145,"x":115,"y":332,"id":3,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":146,"x":731,"y":577,"id":4,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":147,"x":913,"y":316,"id":5,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":148,"x":976,"y":307,"id":30,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":149,"x":275,"y":85,"id":34,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":150,"x":378,"y":130,"id":46,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":151,"x":326,"y":114,"id":49,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":152,"x":725,"y":129,"id":57,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":153,"x":282,"y":191,"id":58,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":154,"x":661,"y":131,"id":60,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":155,"x":216,"y":178,"id":62,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":156,"x":1090,"y":569,"id":65,"time":"2020-06-17 03:27:22.485233","s":12.1963824289},{"index":157,"x":116,"y":317,"id":3,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":158,"x":792,"y":553,"id":4,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":159,"x":977,"y":403,"id":5,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":160,"x":283,"y":231,"id":7,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":161,"x":278,"y":96,"id":34,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":162,"x":373,"y":120,"id":46,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":163,"x":326,"y":132,"id":49,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":164,"x":845,"y":198,"id":57,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":165,"x":677,"y":117,"id":60,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":166,"x":1027,"y":557,"id":65,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":167,"x":85,"y":186,"id":67,"time":"2020-06-17 03:27:23.247507","s":12.9586563307},{"index":168,"x":766,"y":148,"id":1,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":169,"x":112,"y":320,"id":3,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":170,"x":808,"y":511,"id":4,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":171,"x":1172,"y":458,"id":5,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":172,"x":274,"y":230,"id":7,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":173,"x":276,"y":85,"id":34,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":174,"x":373,"y":121,"id":46,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":175,"x":328,"y":117,"id":49,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":176,"x":904,"y":272,"id":57,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":177,"x":676,"y":116,"id":60,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":178,"x":1083,"y":571,"id":65,"time":"2020-06-17 03:27:24.009781","s":13.7209302326},{"index":179,"x":772,"y":141,"id":1,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":180,"x":95,"y":328,"id":3,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":181,"x":812,"y":468,"id":4,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":182,"x":275,"y":234,"id":7,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":183,"x":281,"y":100,"id":34,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":184,"x":371,"y":121,"id":46,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":185,"x":322,"y":134,"id":49,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":186,"x":882,"y":246,"id":57,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":187,"x":661,"y":115,"id":60,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":188,"x":902,"y":586,"id":65,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":189,"x":93,"y":199,"id":67,"time":"2020-06-17 03:27:24.772055","s":14.4832041344},{"index":190,"x":778,"y":134,"id":1,"time":"2020-06-17 03:27:25.534329","s":15.2454780362},{"index":191,"x":838,"y":427,"id":4,"time":"2020-06-17 03:27:25.534329","s":15.2454780362},{"index":192,"x":277,"y":99,"id":34,"time":"2020-06-17 03:27:25.534329","s":15.2454780362},{"index":193,"x":369,"y":121,"id":46,"time":"2020-06-17 03:27:25.534329","s":15.2454780362},{"index":194,"x":319,"y":114,"id":49,"time":"2020-06-17 03:27:25.534329","s":15.2454780362},{"index":195,"x":704,"y":338,"id":57,"time":"2020-06-17 03:27:25.534329","s":15.2454780362},{"index":196,"x":642,"y":116,"id":60,"time":"2020-06-17 03:27:25.534329","s":15.2454780362},{"index":197,"x":815,"y":642,"id":65,"time":"2020-06-17 03:27:25.534329","s":15.2454780362},{"index":198,"x":850,"y":185,"id":79,"time":"2020-06-17 03:27:25.534329","s":15.2454780362},{"index":199,"x":775,"y":138,"id":1,"time":"2020-06-17 03:27:26.296603","s":16.007751938},{"index":200,"x":77,"y":321,"id":3,"time":"2020-06-17 03:27:26.296603","s":16.007751938},{"index":201,"x":843,"y":398,"id":4,"time":"2020-06-17 03:27:26.296603","s":16.007751938},{"index":202,"x":271,"y":184,"id":7,"time":"2020-06-17 03:27:26.296603","s":16.007751938},{"index":203,"x":278,"y":103,"id":34,"time":"2020-06-17 03:27:26.296603","s":16.007751938},{"index":204,"x":370,"y":120,"id":46,"time":"2020-06-17 03:27:26.296603","s":16.007751938},{"index":205,"x":649,"y":117,"id":60,"time":"2020-06-17 03:27:26.296603","s":16.007751938},{"index":206,"x":677,"y":636,"id":65,"time":"2020-06-17 03:27:26.296603","s":16.007751938},{"index":207,"x":850,"y":170,"id":79,"time":"2020-06-17 03:27:26.296603","s":16.007751938},{"index":208,"x":213,"y":149,"id":81,"time":"2020-06-17 03:27:26.296603","s":16.007751938},{"index":209,"x":762,"y":140,"id":1,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":210,"x":80,"y":327,"id":3,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":211,"x":892,"y":351,"id":4,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":212,"x":277,"y":190,"id":7,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":213,"x":371,"y":128,"id":46,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":214,"x":653,"y":116,"id":60,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":215,"x":653,"y":656,"id":65,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":216,"x":93,"y":200,"id":67,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":217,"x":855,"y":203,"id":79,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":218,"x":205,"y":153,"id":81,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":219,"x":269,"y":451,"id":82,"time":"2020-06-17 03:27:27.058877","s":16.7700258398},{"index":220,"x":759,"y":145,"id":1,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":221,"x":73,"y":294,"id":3,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":222,"x":886,"y":302,"id":4,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":223,"x":271,"y":182,"id":7,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":224,"x":322,"y":141,"id":34,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":225,"x":371,"y":120,"id":46,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":226,"x":661,"y":116,"id":60,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":227,"x":88,"y":185,"id":67,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":228,"x":850,"y":179,"id":79,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":229,"x":184,"y":489,"id":82,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":230,"x":693,"y":671,"id":84,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":231,"x":279,"y":109,"id":86,"time":"2020-06-17 03:27:27.821151","s":17.5322997416},{"index":232,"x":760,"y":146,"id":1,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":233,"x":69,"y":321,"id":3,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":234,"x":893,"y":275,"id":4,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":235,"x":373,"y":120,"id":46,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":236,"x":102,"y":197,"id":67,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":237,"x":689,"y":145,"id":76,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":238,"x":830,"y":196,"id":79,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":239,"x":209,"y":538,"id":82,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":240,"x":652,"y":670,"id":84,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":241,"x":280,"y":107,"id":86,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":242,"x":252,"y":143,"id":87,"time":"2020-06-17 03:27:28.583425","s":18.2945736434},{"index":243,"x":757,"y":148,"id":1,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":244,"x":76,"y":349,"id":3,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":245,"x":844,"y":203,"id":4,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":246,"x":253,"y":186,"id":7,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":247,"x":319,"y":114,"id":34,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":248,"x":373,"y":119,"id":46,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":249,"x":105,"y":213,"id":67,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":250,"x":694,"y":146,"id":76,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":251,"x":214,"y":599,"id":82,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":252,"x":654,"y":672,"id":84,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":253,"x":278,"y":106,"id":86,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":254,"x":246,"y":145,"id":87,"time":"2020-06-17 03:27:29.345699","s":19.0568475452},{"index":255,"x":749,"y":149,"id":1,"time":"2020-06-17 03:27:30.107972","s":19.819121447},{"index":256,"x":79,"y":354,"id":3,"time":"2020-06-17 03:27:30.107972","s":19.819121447},{"index":257,"x":820,"y":178,"id":4,"time":"2020-06-17 03:27:30.107972","s":19.819121447},{"index":258,"x":253,"y":198,"id":7,"time":"2020-06-17 03:27:30.107972","s":19.819121447},{"index":259,"x":273,"y":105,"id":34,"time":"2020-06-17 03:27:30.107972","s":19.819121447},{"index":260,"x":372,"y":121,"id":46,"time":"2020-06-17 03:27:30.107972","s":19.819121447},{"index":261,"x":108,"y":208,"id":67,"time":"2020-06-17 03:27:30.107972","s":19.819121447},{"index":262,"x":212,"y":592,"id":82,"time":"2020-06-17 03:27:30.107972","s":19.819121447},{"index":263,"x":285,"y":138,"id":86,"time":"2020-06-17 03:27:30.107972","s":19.819121447},{"index":264,"x":245,"y":145,"id":87,"time":"2020-06-17 03:27:30.107972","s":19.819121447},{"index":265,"x":64,"y":347,"id":3,"time":"2020-06-17 03:27:30.870246","s":20.5813953488},{"index":266,"x":788,"y":173,"id":4,"time":"2020-06-17 03:27:30.870246","s":20.5813953488},{"index":267,"x":253,"y":223,"id":7,"time":"2020-06-17 03:27:30.870246","s":20.5813953488},{"index":268,"x":372,"y":122,"id":46,"time":"2020-06-17 03:27:30.870246","s":20.5813953488},{"index":269,"x":105,"y":221,"id":67,"time":"2020-06-17 03:27:30.870246","s":20.5813953488},{"index":270,"x":219,"y":493,"id":82,"time":"2020-06-17 03:27:30.870246","s":20.5813953488},{"index":271,"x":286,"y":137,"id":86,"time":"2020-06-17 03:27:30.870246","s":20.5813953488},{"index":272,"x":248,"y":135,"id":87,"time":"2020-06-17 03:27:30.870246","s":20.5813953488},{"index":273,"x":69,"y":280,"id":3,"time":"2020-06-17 03:27:31.632520","s":21.3436692506},{"index":274,"x":804,"y":175,"id":4,"time":"2020-06-17 03:27:31.632520","s":21.3436692506},{"index":275,"x":275,"y":180,"id":7,"time":"2020-06-17 03:27:31.632520","s":21.3436692506},{"index":276,"x":277,"y":102,"id":34,"time":"2020-06-17 03:27:31.632520","s":21.3436692506},{"index":277,"x":372,"y":121,"id":46,"time":"2020-06-17 03:27:31.632520","s":21.3436692506},{"index":278,"x":108,"y":203,"id":67,"time":"2020-06-17 03:27:31.632520","s":21.3436692506},{"index":279,"x":193,"y":489,"id":82,"time":"2020-06-17 03:27:31.632520","s":21.3436692506},{"index":280,"x":680,"y":676,"id":84,"time":"2020-06-17 03:27:31.632520","s":21.3436692506},{"index":281,"x":287,"y":243,"id":91,"time":"2020-06-17 03:27:31.632520","s":21.3436692506},{"index":282,"x":750,"y":107,"id":1,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":283,"x":78,"y":334,"id":3,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":284,"x":793,"y":172,"id":4,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":285,"x":318,"y":110,"id":34,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":286,"x":372,"y":121,"id":46,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":287,"x":122,"y":209,"id":67,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":288,"x":228,"y":471,"id":82,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":289,"x":272,"y":85,"id":86,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":290,"x":300,"y":240,"id":91,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":291,"x":647,"y":111,"id":92,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":292,"x":227,"y":149,"id":93,"time":"2020-06-17 03:27:32.394794","s":22.1059431525},{"index":293,"x":84,"y":343,"id":3,"time":"2020-06-17 03:27:33.157068","s":22.8682170543},{"index":294,"x":791,"y":175,"id":4,"time":"2020-06-17 03:27:33.157068","s":22.8682170543},{"index":295,"x":304,"y":205,"id":7,"time":"2020-06-17 03:27:33.157068","s":22.8682170543},{"index":296,"x":318,"y":114,"id":34,"time":"2020-06-17 03:27:33.157068","s":22.8682170543},{"index":297,"x":372,"y":120,"id":46,"time":"2020-06-17 03:27:33.157068","s":22.8682170543},{"index":298,"x":94,"y":207,"id":67,"time":"2020-06-17 03:27:33.157068","s":22.8682170543},{"index":299,"x":292,"y":442,"id":82,"time":"2020-06-17 03:27:33.157068","s":22.8682170543},{"index":300,"x":273,"y":87,"id":86,"time":"2020-06-17 03:27:33.157068","s":22.8682170543},{"index":301,"x":217,"y":168,"id":93,"time":"2020-06-17 03:27:33.157068","s":22.8682170543},{"index":302,"x":744,"y":101,"id":1,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":303,"x":85,"y":342,"id":3,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":304,"x":779,"y":179,"id":4,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":305,"x":319,"y":115,"id":34,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":306,"x":373,"y":113,"id":46,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":307,"x":107,"y":211,"id":67,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":308,"x":401,"y":356,"id":82,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":309,"x":272,"y":90,"id":86,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":310,"x":289,"y":212,"id":91,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":311,"x":214,"y":170,"id":93,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":312,"x":688,"y":143,"id":94,"time":"2020-06-17 03:27:33.919342","s":23.6304909561},{"index":313,"x":746,"y":101,"id":1,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":314,"x":100,"y":331,"id":3,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":315,"x":778,"y":175,"id":4,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":316,"x":320,"y":112,"id":34,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":317,"x":373,"y":121,"id":46,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":318,"x":436,"y":299,"id":82,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":319,"x":271,"y":99,"id":86,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":320,"x":293,"y":244,"id":91,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":321,"x":656,"y":105,"id":92,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":322,"x":215,"y":166,"id":93,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":323,"x":689,"y":144,"id":94,"time":"2020-06-17 03:27:34.681616","s":24.3927648579},{"index":324,"x":747,"y":105,"id":1,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":325,"x":106,"y":317,"id":3,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":326,"x":772,"y":167,"id":4,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":327,"x":321,"y":109,"id":34,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":328,"x":372,"y":120,"id":46,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":329,"x":116,"y":201,"id":67,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":330,"x":460,"y":234,"id":82,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":331,"x":274,"y":85,"id":86,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":332,"x":301,"y":239,"id":91,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":333,"x":639,"y":106,"id":92,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":334,"x":229,"y":166,"id":93,"time":"2020-06-17 03:27:35.443890","s":25.1550387597},{"index":335,"x":100,"y":319,"id":3,"time":"2020-06-17 03:27:36.206164","s":25.9173126615},{"index":336,"x":761,"y":167,"id":4,"time":"2020-06-17 03:27:36.206164","s":25.9173126615},{"index":337,"x":312,"y":116,"id":34,"time":"2020-06-17 03:27:36.206164","s":25.9173126615},{"index":338,"x":371,"y":121,"id":46,"time":"2020-06-17 03:27:36.206164","s":25.9173126615},{"index":339,"x":122,"y":204,"id":67,"time":"2020-06-17 03:27:36.206164","s":25.9173126615},{"index":340,"x":431,"y":216,"id":82,"time":"2020-06-17 03:27:36.206164","s":25.9173126615},{"index":341,"x":255,"y":85,"id":86,"time":"2020-06-17 03:27:36.206164","s":25.9173126615},{"index":342,"x":280,"y":236,"id":91,"time":"2020-06-17 03:27:36.206164","s":25.9173126615},{"index":343,"x":630,"y":108,"id":92,"time":"2020-06-17 03:27:36.206164","s":25.9173126615},{"index":344,"x":826,"y":182,"id":96,"time":"2020-06-17 03:27:36.206164","s":25.9173126615},{"index":345,"x":97,"y":326,"id":3,"time":"2020-06-17 03:27:36.968438","s":26.6795865633},{"index":346,"x":722,"y":150,"id":4,"time":"2020-06-17 03:27:36.968438","s":26.6795865633},{"index":347,"x":315,"y":110,"id":34,"time":"2020-06-17 03:27:36.968438","s":26.6795865633},{"index":348,"x":119,"y":210,"id":67,"time":"2020-06-17 03:27:36.968438","s":26.6795865633},{"index":349,"x":382,"y":174,"id":82,"time":"2020-06-17 03:27:36.968438","s":26.6795865633},{"index":350,"x":271,"y":83,"id":86,"time":"2020-06-17 03:27:36.968438","s":26.6795865633},{"index":351,"x":288,"y":242,"id":91,"time":"2020-06-17 03:27:36.968438","s":26.6795865633},{"index":352,"x":653,"y":109,"id":92,"time":"2020-06-17 03:27:36.968438","s":26.6795865633},{"index":353,"x":829,"y":194,"id":96,"time":"2020-06-17 03:27:36.968438","s":26.6795865633},{"index":354,"x":692,"y":122,"id":1,"time":"2020-06-17 03:27:37.730711","s":27.4418604651},{"index":355,"x":102,"y":318,"id":3,"time":"2020-06-17 03:27:37.730711","s":27.4418604651},{"index":356,"x":752,"y":157,"id":4,"time":"2020-06-17 03:27:37.730711","s":27.4418604651},{"index":357,"x":310,"y":99,"id":34,"time":"2020-06-17 03:27:37.730711","s":27.4418604651},{"index":358,"x":120,"y":211,"id":67,"time":"2020-06-17 03:27:37.730711","s":27.4418604651},{"index":359,"x":369,"y":147,"id":82,"time":"2020-06-17 03:27:37.730711","s":27.4418604651},{"index":360,"x":258,"y":85,"id":86,"time":"2020-06-17 03:27:37.730711","s":27.4418604651},{"index":361,"x":315,"y":247,"id":91,"time":"2020-06-17 03:27:37.730711","s":27.4418604651},{"index":362,"x":825,"y":186,"id":96,"time":"2020-06-17 03:27:37.730711","s":27.4418604651},{"index":363,"x":242,"y":169,"id":99,"time":"2020-06-17 03:27:37.730711","s":27.4418604651},{"index":364,"x":705,"y":131,"id":1,"time":"2020-06-17 03:27:38.492985","s":28.2041343669},{"index":365,"x":101,"y":318,"id":3,"time":"2020-06-17 03:27:38.492985","s":28.2041343669},{"index":366,"x":745,"y":157,"id":4,"time":"2020-06-17 03:27:38.492985","s":28.2041343669},{"index":367,"x":305,"y":86,"id":34,"time":"2020-06-17 03:27:38.492985","s":28.2041343669},{"index":368,"x":123,"y":205,"id":67,"time":"2020-06-17 03:27:38.492985","s":28.2041343669},{"index":369,"x":384,"y":134,"id":82,"time":"2020-06-17 03:27:38.492985","s":28.2041343669},{"index":370,"x":255,"y":85,"id":86,"time":"2020-06-17 03:27:38.492985","s":28.2041343669},{"index":371,"x":314,"y":246,"id":91,"time":"2020-06-17 03:27:38.492985","s":28.2041343669},{"index":372,"x":820,"y":191,"id":96,"time":"2020-06-17 03:27:38.492985","s":28.2041343669},{"index":373,"x":243,"y":151,"id":99,"time":"2020-06-17 03:27:38.492985","s":28.2041343669},{"index":374,"x":104,"y":318,"id":3,"time":"2020-06-17 03:27:39.255259","s":28.9664082687},{"index":375,"x":713,"y":147,"id":4,"time":"2020-06-17 03:27:39.255259","s":28.9664082687},{"index":376,"x":294,"y":84,"id":34,"time":"2020-06-17 03:27:39.255259","s":28.9664082687},{"index":377,"x":125,"y":207,"id":67,"time":"2020-06-17 03:27:39.255259","s":28.9664082687},{"index":378,"x":253,"y":84,"id":86,"time":"2020-06-17 03:27:39.255259","s":28.9664082687},{"index":379,"x":319,"y":246,"id":91,"time":"2020-06-17 03:27:39.255259","s":28.9664082687},{"index":380,"x":819,"y":197,"id":96,"time":"2020-06-17 03:27:39.255259","s":28.9664082687},{"index":381,"x":244,"y":174,"id":99,"time":"2020-06-17 03:27:39.255259","s":28.9664082687},{"index":382,"x":344,"y":101,"id":101,"time":"2020-06-17 03:27:39.255259","s":28.9664082687},{"index":383,"x":471,"y":68,"id":103,"time":"2020-06-17 03:27:39.255259","s":28.9664082687},{"index":384,"x":100,"y":319,"id":3,"time":"2020-06-17 03:27:40.017533","s":29.7286821705},{"index":385,"x":711,"y":149,"id":4,"time":"2020-06-17 03:27:40.017533","s":29.7286821705},{"index":386,"x":126,"y":213,"id":67,"time":"2020-06-17 03:27:40.017533","s":29.7286821705},{"index":387,"x":250,"y":78,"id":86,"time":"2020-06-17 03:27:40.017533","s":29.7286821705},{"index":388,"x":292,"y":243,"id":91,"time":"2020-06-17 03:27:40.017533","s":29.7286821705},{"index":389,"x":818,"y":198,"id":96,"time":"2020-06-17 03:27:40.017533","s":29.7286821705},{"index":390,"x":242,"y":133,"id":99,"time":"2020-06-17 03:27:40.017533","s":29.7286821705},{"index":391,"x":348,"y":132,"id":101,"time":"2020-06-17 03:27:40.017533","s":29.7286821705},{"index":392,"x":512,"y":56,"id":103,"time":"2020-06-17 03:27:40.017533","s":29.7286821705},{"index":393,"x":661,"y":113,"id":104,"time":"2020-06-17 03:27:40.017533","s":29.7286821705},{"index":394,"x":101,"y":317,"id":3,"time":"2020-06-17 03:27:40.779807","s":30.4909560724},{"index":395,"x":713,"y":143,"id":4,"time":"2020-06-17 03:27:40.779807","s":30.4909560724},{"index":396,"x":120,"y":206,"id":67,"time":"2020-06-17 03:27:40.779807","s":30.4909560724},{"index":397,"x":296,"y":249,"id":91,"time":"2020-06-17 03:27:40.779807","s":30.4909560724},{"index":398,"x":820,"y":198,"id":96,"time":"2020-06-17 03:27:40.779807","s":30.4909560724},{"index":399,"x":349,"y":132,"id":101,"time":"2020-06-17 03:27:40.779807","s":30.4909560724},{"index":400,"x":488,"y":48,"id":103,"time":"2020-06-17 03:27:40.779807","s":30.4909560724},{"index":401,"x":659,"y":108,"id":104,"time":"2020-06-17 03:27:40.779807","s":30.4909560724},{"index":402,"x":305,"y":101,"id":105,"time":"2020-06-17 03:27:40.779807","s":30.4909560724},{"index":403,"x":88,"y":318,"id":3,"time":"2020-06-17 03:27:41.542081","s":31.2532299742},{"index":404,"x":715,"y":153,"id":4,"time":"2020-06-17 03:27:41.542081","s":31.2532299742},{"index":405,"x":306,"y":87,"id":34,"time":"2020-06-17 03:27:41.542081","s":31.2532299742},{"index":406,"x":258,"y":81,"id":86,"time":"2020-06-17 03:27:41.542081","s":31.2532299742},{"index":407,"x":308,"y":247,"id":91,"time":"2020-06-17 03:27:41.542081","s":31.2532299742},{"index":408,"x":815,"y":198,"id":96,"time":"2020-06-17 03:27:41.542081","s":31.2532299742},{"index":409,"x":256,"y":140,"id":99,"time":"2020-06-17 03:27:41.542081","s":31.2532299742},{"index":410,"x":351,"y":97,"id":101,"time":"2020-06-17 03:27:41.542081","s":31.2532299742},{"index":411,"x":88,"y":314,"id":3,"time":"2020-06-17 03:27:42.304355","s":32.015503876},{"index":412,"x":306,"y":85,"id":34,"time":"2020-06-17 03:27:42.304355","s":32.015503876},{"index":413,"x":257,"y":78,"id":86,"time":"2020-06-17 03:27:42.304355","s":32.015503876},{"index":414,"x":310,"y":250,"id":91,"time":"2020-06-17 03:27:42.304355","s":32.015503876},{"index":415,"x":823,"y":196,"id":96,"time":"2020-06-17 03:27:42.304355","s":32.015503876},{"index":416,"x":250,"y":138,"id":99,"time":"2020-06-17 03:27:42.304355","s":32.015503876},{"index":417,"x":349,"y":99,"id":101,"time":"2020-06-17 03:27:42.304355","s":32.015503876},{"index":418,"x":92,"y":308,"id":3,"time":"2020-06-17 03:27:43.066629","s":32.7777777778},{"index":419,"x":367,"y":267,"id":91,"time":"2020-06-17 03:27:43.066629","s":32.7777777778},{"index":420,"x":823,"y":197,"id":96,"time":"2020-06-17 03:27:43.066629","s":32.7777777778},{"index":421,"x":261,"y":197,"id":99,"time":"2020-06-17 03:27:43.066629","s":32.7777777778},{"index":422,"x":351,"y":87,"id":101,"time":"2020-06-17 03:27:43.066629","s":32.7777777778},{"index":423,"x":306,"y":113,"id":105,"time":"2020-06-17 03:27:43.066629","s":32.7777777778},{"index":424,"x":657,"y":671,"id":112,"time":"2020-06-17 03:27:43.066629","s":32.7777777778},{"index":425,"x":124,"y":266,"id":113,"time":"2020-06-17 03:27:43.066629","s":32.7777777778},{"index":426,"x":94,"y":316,"id":3,"time":"2020-06-17 03:27:43.828903","s":33.5400516796},{"index":427,"x":105,"y":207,"id":67,"time":"2020-06-17 03:27:43.828903","s":33.5400516796},{"index":428,"x":410,"y":276,"id":91,"time":"2020-06-17 03:27:43.828903","s":33.5400516796},{"index":429,"x":823,"y":196,"id":96,"time":"2020-06-17 03:27:43.828903","s":33.5400516796},{"index":430,"x":261,"y":205,"id":99,"time":"2020-06-17 03:27:43.828903","s":33.5400516796},{"index":431,"x":309,"y":141,"id":105,"time":"2020-06-17 03:27:43.828903","s":33.5400516796},{"index":432,"x":656,"y":671,"id":112,"time":"2020-06-17 03:27:43.828903","s":33.5400516796},{"index":433,"x":348,"y":139,"id":114,"time":"2020-06-17 03:27:43.828903","s":33.5400516796},{"index":434,"x":96,"y":314,"id":3,"time":"2020-06-17 03:27:44.591177","s":34.3023255814},{"index":435,"x":731,"y":147,"id":4,"time":"2020-06-17 03:27:44.591177","s":34.3023255814},{"index":436,"x":94,"y":204,"id":67,"time":"2020-06-17 03:27:44.591177","s":34.3023255814},{"index":437,"x":423,"y":268,"id":91,"time":"2020-06-17 03:27:44.591177","s":34.3023255814},{"index":438,"x":822,"y":198,"id":96,"time":"2020-06-17 03:27:44.591177","s":34.3023255814},{"index":439,"x":271,"y":204,"id":99,"time":"2020-06-17 03:27:44.591177","s":34.3023255814},{"index":440,"x":309,"y":139,"id":105,"time":"2020-06-17 03:27:44.591177","s":34.3023255814},{"index":441,"x":671,"y":99,"id":108,"time":"2020-06-17 03:27:44.591177","s":34.3023255814},{"index":442,"x":657,"y":671,"id":112,"time":"2020-06-17 03:27:44.591177","s":34.3023255814},{"index":443,"x":347,"y":136,"id":114,"time":"2020-06-17 03:27:44.591177","s":34.3023255814},{"index":444,"x":97,"y":311,"id":3,"time":"2020-06-17 03:27:45.353450","s":35.0645994832},{"index":445,"x":738,"y":143,"id":4,"time":"2020-06-17 03:27:45.353450","s":35.0645994832},{"index":446,"x":95,"y":207,"id":67,"time":"2020-06-17 03:27:45.353450","s":35.0645994832},{"index":447,"x":476,"y":281,"id":91,"time":"2020-06-17 03:27:45.353450","s":35.0645994832},{"index":448,"x":822,"y":197,"id":96,"time":"2020-06-17 03:27:45.353450","s":35.0645994832},{"index":449,"x":243,"y":168,"id":99,"time":"2020-06-17 03:27:45.353450","s":35.0645994832},{"index":450,"x":667,"y":101,"id":108,"time":"2020-06-17 03:27:45.353450","s":35.0645994832},{"index":451,"x":661,"y":672,"id":112,"time":"2020-06-17 03:27:45.353450","s":35.0645994832},{"index":452,"x":349,"y":134,"id":114,"time":"2020-06-17 03:27:45.353450","s":35.0645994832},{"index":453,"x":95,"y":289,"id":3,"time":"2020-06-17 03:27:46.115724","s":35.826873385},{"index":454,"x":738,"y":140,"id":4,"time":"2020-06-17 03:27:46.115724","s":35.826873385},{"index":455,"x":96,"y":212,"id":67,"time":"2020-06-17 03:27:46.115724","s":35.826873385},{"index":456,"x":824,"y":196,"id":96,"time":"2020-06-17 03:27:46.115724","s":35.826873385},{"index":457,"x":252,"y":172,"id":99,"time":"2020-06-17 03:27:46.115724","s":35.826873385},{"index":458,"x":261,"y":101,"id":105,"time":"2020-06-17 03:27:46.115724","s":35.826873385},{"index":459,"x":662,"y":101,"id":108,"time":"2020-06-17 03:27:46.115724","s":35.826873385},{"index":460,"x":661,"y":672,"id":112,"time":"2020-06-17 03:27:46.115724","s":35.826873385},{"index":461,"x":349,"y":134,"id":114,"time":"2020-06-17 03:27:46.115724","s":35.826873385},{"index":462,"x":632,"y":364,"id":115,"time":"2020-06-17 03:27:46.115724","s":35.826873385},{"index":463,"x":83,"y":327,"id":3,"time":"2020-06-17 03:27:46.877998","s":36.5891472868},{"index":464,"x":739,"y":149,"id":4,"time":"2020-06-17 03:27:46.877998","s":36.5891472868},{"index":465,"x":822,"y":197,"id":96,"time":"2020-06-17 03:27:46.877998","s":36.5891472868},{"index":466,"x":247,"y":184,"id":99,"time":"2020-06-17 03:27:46.877998","s":36.5891472868},{"index":467,"x":643,"y":108,"id":108,"time":"2020-06-17 03:27:46.877998","s":36.5891472868},{"index":468,"x":654,"y":666,"id":112,"time":"2020-06-17 03:27:46.877998","s":36.5891472868},{"index":469,"x":345,"y":134,"id":114,"time":"2020-06-17 03:27:46.877998","s":36.5891472868},{"index":470,"x":811,"y":551,"id":115,"time":"2020-06-17 03:27:46.877998","s":36.5891472868},{"index":471,"x":76,"y":336,"id":3,"time":"2020-06-17 03:27:47.640272","s":37.3514211886},{"index":472,"x":738,"y":146,"id":4,"time":"2020-06-17 03:27:47.640272","s":37.3514211886},{"index":473,"x":113,"y":217,"id":67,"time":"2020-06-17 03:27:47.640272","s":37.3514211886},{"index":474,"x":824,"y":196,"id":96,"time":"2020-06-17 03:27:47.640272","s":37.3514211886},{"index":475,"x":239,"y":168,"id":99,"time":"2020-06-17 03:27:47.640272","s":37.3514211886},{"index":476,"x":648,"y":104,"id":108,"time":"2020-06-17 03:27:47.640272","s":37.3514211886},{"index":477,"x":612,"y":639,"id":112,"time":"2020-06-17 03:27:47.640272","s":37.3514211886},{"index":478,"x":347,"y":134,"id":114,"time":"2020-06-17 03:27:47.640272","s":37.3514211886},{"index":479,"x":943,"y":642,"id":115,"time":"2020-06-17 03:27:47.640272","s":37.3514211886},{"index":480,"x":112,"y":240,"id":116,"time":"2020-06-17 03:27:47.640272","s":37.3514211886},{"index":481,"x":105,"y":331,"id":3,"time":"2020-06-17 03:27:48.402546","s":38.1136950904},{"index":482,"x":742,"y":147,"id":4,"time":"2020-06-17 03:27:48.402546","s":38.1136950904},{"index":483,"x":823,"y":197,"id":96,"time":"2020-06-17 03:27:48.402546","s":38.1136950904},{"index":484,"x":245,"y":166,"id":99,"time":"2020-06-17 03:27:48.402546","s":38.1136950904},{"index":485,"x":261,"y":83,"id":105,"time":"2020-06-17 03:27:48.402546","s":38.1136950904},{"index":486,"x":648,"y":119,"id":108,"time":"2020-06-17 03:27:48.402546","s":38.1136950904},{"index":487,"x":532,"y":674,"id":112,"time":"2020-06-17 03:27:48.402546","s":38.1136950904},{"index":488,"x":347,"y":134,"id":114,"time":"2020-06-17 03:27:48.402546","s":38.1136950904},{"index":489,"x":104,"y":332,"id":3,"time":"2020-06-17 03:27:49.164820","s":38.8759689922},{"index":490,"x":717,"y":142,"id":4,"time":"2020-06-17 03:27:49.164820","s":38.8759689922},{"index":491,"x":821,"y":197,"id":96,"time":"2020-06-17 03:27:49.164820","s":38.8759689922},{"index":492,"x":296,"y":171,"id":99,"time":"2020-06-17 03:27:49.164820","s":38.8759689922},{"index":493,"x":646,"y":104,"id":108,"time":"2020-06-17 03:27:49.164820","s":38.8759689922},{"index":494,"x":344,"y":133,"id":114,"time":"2020-06-17 03:27:49.164820","s":38.8759689922},{"index":495,"x":247,"y":143,"id":120,"time":"2020-06-17 03:27:49.164820","s":38.8759689922},{"index":496,"x":218,"y":185,"id":121,"time":"2020-06-17 03:27:49.164820","s":38.8759689922},{"index":497,"x":78,"y":333,"id":3,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":498,"x":740,"y":148,"id":4,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":499,"x":819,"y":198,"id":96,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":500,"x":295,"y":179,"id":99,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":501,"x":648,"y":109,"id":108,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":502,"x":345,"y":135,"id":114,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":503,"x":109,"y":236,"id":116,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":504,"x":243,"y":143,"id":120,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":505,"x":220,"y":186,"id":121,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":506,"x":491,"y":676,"id":123,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":507,"x":948,"y":607,"id":124,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":508,"x":323,"y":264,"id":125,"time":"2020-06-17 03:27:49.927094","s":39.6382428941},{"index":509,"x":74,"y":338,"id":3,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":510,"x":719,"y":149,"id":4,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":511,"x":820,"y":197,"id":96,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":512,"x":295,"y":204,"id":99,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":513,"x":647,"y":108,"id":108,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":514,"x":346,"y":135,"id":114,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":515,"x":104,"y":240,"id":116,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":516,"x":247,"y":147,"id":120,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":517,"x":219,"y":186,"id":121,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":518,"x":522,"y":665,"id":123,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":519,"x":965,"y":577,"id":124,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":520,"x":342,"y":267,"id":125,"time":"2020-06-17 03:27:50.689368","s":40.4005167959},{"index":521,"x":76,"y":331,"id":3,"time":"2020-06-17 03:27:51.451642","s":41.1627906977},{"index":522,"x":709,"y":147,"id":4,"time":"2020-06-17 03:27:51.451642","s":41.1627906977},{"index":523,"x":821,"y":197,"id":96,"time":"2020-06-17 03:27:51.451642","s":41.1627906977},{"index":524,"x":270,"y":207,"id":99,"time":"2020-06-17 03:27:51.451642","s":41.1627906977},{"index":525,"x":640,"y":107,"id":108,"time":"2020-06-17 03:27:51.451642","s":41.1627906977},{"index":526,"x":99,"y":234,"id":116,"time":"2020-06-17 03:27:51.451642","s":41.1627906977},{"index":527,"x":520,"y":663,"id":123,"time":"2020-06-17 03:27:51.451642","s":41.1627906977},{"index":528,"x":898,"y":523,"id":124,"time":"2020-06-17 03:27:51.451642","s":41.1627906977},{"index":529,"x":76,"y":330,"id":3,"time":"2020-06-17 03:27:52.213916","s":41.9250645995},{"index":530,"x":716,"y":132,"id":4,"time":"2020-06-17 03:27:52.213916","s":41.9250645995},{"index":531,"x":819,"y":198,"id":96,"time":"2020-06-17 03:27:52.213916","s":41.9250645995},{"index":532,"x":259,"y":222,"id":99,"time":"2020-06-17 03:27:52.213916","s":41.9250645995},{"index":533,"x":639,"y":115,"id":108,"time":"2020-06-17 03:27:52.213916","s":41.9250645995},{"index":534,"x":348,"y":135,"id":114,"time":"2020-06-17 03:27:52.213916","s":41.9250645995},{"index":535,"x":91,"y":233,"id":116,"time":"2020-06-17 03:27:52.213916","s":41.9250645995},{"index":536,"x":515,"y":653,"id":123,"time":"2020-06-17 03:27:52.213916","s":41.9250645995},{"index":537,"x":738,"y":418,"id":124,"time":"2020-06-17 03:27:52.213916","s":41.9250645995},{"index":538,"x":340,"y":270,"id":125,"time":"2020-06-17 03:27:52.213916","s":41.9250645995},{"index":539,"x":61,"y":346,"id":3,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":540,"x":729,"y":141,"id":4,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":541,"x":819,"y":198,"id":96,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":542,"x":293,"y":235,"id":99,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":543,"x":646,"y":117,"id":108,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":544,"x":347,"y":135,"id":114,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":545,"x":92,"y":234,"id":116,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":546,"x":224,"y":187,"id":121,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":547,"x":511,"y":650,"id":123,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":548,"x":588,"y":288,"id":124,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":549,"x":342,"y":269,"id":125,"time":"2020-06-17 03:27:52.976190","s":42.6873385013},{"index":550,"x":72,"y":349,"id":3,"time":"2020-06-17 03:27:53.738463","s":43.4496124031},{"index":551,"x":713,"y":147,"id":4,"time":"2020-06-17 03:27:53.738463","s":43.4496124031},{"index":552,"x":819,"y":197,"id":96,"time":"2020-06-17 03:27:53.738463","s":43.4496124031},{"index":553,"x":631,"y":111,"id":108,"time":"2020-06-17 03:27:53.738463","s":43.4496124031},{"index":554,"x":346,"y":134,"id":114,"time":"2020-06-17 03:27:53.738463","s":43.4496124031},{"index":555,"x":92,"y":235,"id":116,"time":"2020-06-17 03:27:53.738463","s":43.4496124031},{"index":556,"x":506,"y":647,"id":123,"time":"2020-06-17 03:27:53.738463","s":43.4496124031},{"index":557,"x":457,"y":264,"id":124,"time":"2020-06-17 03:27:53.738463","s":43.4496124031},{"index":558,"x":340,"y":266,"id":125,"time":"2020-06-17 03:27:53.738463","s":43.4496124031},{"index":559,"x":78,"y":372,"id":3,"time":"2020-06-17 03:27:54.500737","s":44.2118863049},{"index":560,"x":708,"y":143,"id":4,"time":"2020-06-17 03:27:54.500737","s":44.2118863049},{"index":561,"x":821,"y":198,"id":96,"time":"2020-06-17 03:27:54.500737","s":44.2118863049},{"index":562,"x":286,"y":233,"id":99,"time":"2020-06-17 03:27:54.500737","s":44.2118863049},{"index":563,"x":627,"y":109,"id":108,"time":"2020-06-17 03:27:54.500737","s":44.2118863049},{"index":564,"x":347,"y":113,"id":114,"time":"2020-06-17 03:27:54.500737","s":44.2118863049},{"index":565,"x":98,"y":241,"id":116,"time":"2020-06-17 03:27:54.500737","s":44.2118863049},{"index":566,"x":513,"y":651,"id":123,"time":"2020-06-17 03:27:54.500737","s":44.2118863049},{"index":567,"x":406,"y":248,"id":124,"time":"2020-06-17 03:27:54.500737","s":44.2118863049},{"index":568,"x":256,"y":114,"id":130,"time":"2020-06-17 03:27:54.500737","s":44.2118863049},{"index":569,"x":82,"y":364,"id":3,"time":"2020-06-17 03:27:55.263011","s":44.9741602067},{"index":570,"x":719,"y":142,"id":4,"time":"2020-06-17 03:27:55.263011","s":44.9741602067},{"index":571,"x":821,"y":197,"id":96,"time":"2020-06-17 03:27:55.263011","s":44.9741602067},{"index":572,"x":286,"y":213,"id":99,"time":"2020-06-17 03:27:55.263011","s":44.9741602067},{"index":573,"x":629,"y":110,"id":108,"time":"2020-06-17 03:27:55.263011","s":44.9741602067},{"index":574,"x":110,"y":250,"id":116,"time":"2020-06-17 03:27:55.263011","s":44.9741602067},{"index":575,"x":524,"y":654,"id":123,"time":"2020-06-17 03:27:55.263011","s":44.9741602067},{"index":576,"x":380,"y":236,"id":124,"time":"2020-06-17 03:27:55.263011","s":44.9741602067},{"index":577,"x":253,"y":112,"id":130,"time":"2020-06-17 03:27:55.263011","s":44.9741602067},{"index":578,"x":81,"y":348,"id":3,"time":"2020-06-17 03:27:56.025285","s":45.7364341085},{"index":579,"x":820,"y":198,"id":96,"time":"2020-06-17 03:27:56.025285","s":45.7364341085},{"index":580,"x":260,"y":211,"id":99,"time":"2020-06-17 03:27:56.025285","s":45.7364341085},{"index":581,"x":637,"y":112,"id":108,"time":"2020-06-17 03:27:56.025285","s":45.7364341085},{"index":582,"x":348,"y":111,"id":114,"time":"2020-06-17 03:27:56.025285","s":45.7364341085},{"index":583,"x":120,"y":244,"id":116,"time":"2020-06-17 03:27:56.025285","s":45.7364341085},{"index":584,"x":544,"y":647,"id":123,"time":"2020-06-17 03:27:56.025285","s":45.7364341085},{"index":585,"x":392,"y":244,"id":124,"time":"2020-06-17 03:27:56.025285","s":45.7364341085},{"index":586,"x":74,"y":379,"id":3,"time":"2020-06-17 03:27:56.787559","s":46.4987080103},{"index":587,"x":805,"y":174,"id":4,"time":"2020-06-17 03:27:56.787559","s":46.4987080103},{"index":588,"x":259,"y":213,"id":99,"time":"2020-06-17 03:27:56.787559","s":46.4987080103},{"index":589,"x":638,"y":111,"id":108,"time":"2020-06-17 03:27:56.787559","s":46.4987080103},{"index":590,"x":347,"y":130,"id":114,"time":"2020-06-17 03:27:56.787559","s":46.4987080103},{"index":591,"x":106,"y":246,"id":116,"time":"2020-06-17 03:27:56.787559","s":46.4987080103},{"index":592,"x":547,"y":634,"id":123,"time":"2020-06-17 03:27:56.787559","s":46.4987080103},{"index":593,"x":440,"y":272,"id":124,"time":"2020-06-17 03:27:56.787559","s":46.4987080103},{"index":594,"x":343,"y":267,"id":125,"time":"2020-06-17 03:27:56.787559","s":46.4987080103},{"index":595,"x":139,"y":429,"id":3,"time":"2020-06-17 03:27:57.549833","s":47.2609819121},{"index":596,"x":854,"y":219,"id":4,"time":"2020-06-17 03:27:57.549833","s":47.2609819121},{"index":597,"x":257,"y":216,"id":99,"time":"2020-06-17 03:27:57.549833","s":47.2609819121},{"index":598,"x":629,"y":110,"id":108,"time":"2020-06-17 03:27:57.549833","s":47.2609819121},{"index":599,"x":347,"y":135,"id":114,"time":"2020-06-17 03:27:57.549833","s":47.2609819121},{"index":600,"x":549,"y":616,"id":123,"time":"2020-06-17 03:27:57.549833","s":47.2609819121},{"index":601,"x":322,"y":276,"id":125,"time":"2020-06-17 03:27:57.549833","s":47.2609819121},{"index":602,"x":707,"y":131,"id":135,"time":"2020-06-17 03:27:57.549833","s":47.2609819121},{"index":603,"x":179,"y":448,"id":3,"time":"2020-06-17 03:27:58.312107","s":48.023255814},{"index":604,"x":915,"y":292,"id":4,"time":"2020-06-17 03:27:58.312107","s":48.023255814},{"index":605,"x":811,"y":195,"id":96,"time":"2020-06-17 03:27:58.312107","s":48.023255814},{"index":606,"x":260,"y":201,"id":99,"time":"2020-06-17 03:27:58.312107","s":48.023255814},{"index":607,"x":639,"y":112,"id":108,"time":"2020-06-17 03:27:58.312107","s":48.023255814},{"index":608,"x":539,"y":613,"id":123,"time":"2020-06-17 03:27:58.312107","s":48.023255814},{"index":609,"x":451,"y":347,"id":124,"time":"2020-06-17 03:27:58.312107","s":48.023255814},{"index":610,"x":319,"y":289,"id":125,"time":"2020-06-17 03:27:58.312107","s":48.023255814},{"index":611,"x":94,"y":234,"id":137,"time":"2020-06-17 03:27:58.312107","s":48.023255814},{"index":612,"x":249,"y":473,"id":3,"time":"2020-06-17 03:27:59.074381","s":48.7855297158},{"index":613,"x":908,"y":342,"id":4,"time":"2020-06-17 03:27:59.074381","s":48.7855297158},{"index":614,"x":793,"y":188,"id":96,"time":"2020-06-17 03:27:59.074381","s":48.7855297158},{"index":615,"x":243,"y":211,"id":99,"time":"2020-06-17 03:27:59.074381","s":48.7855297158},{"index":616,"x":640,"y":111,"id":108,"time":"2020-06-17 03:27:59.074381","s":48.7855297158},{"index":617,"x":523,"y":614,"id":123,"time":"2020-06-17 03:27:59.074381","s":48.7855297158},{"index":618,"x":469,"y":296,"id":124,"time":"2020-06-17 03:27:59.074381","s":48.7855297158},{"index":619,"x":322,"y":286,"id":125,"time":"2020-06-17 03:27:59.074381","s":48.7855297158},{"index":620,"x":106,"y":245,"id":137,"time":"2020-06-17 03:27:59.074381","s":48.7855297158},{"index":621,"x":718,"y":152,"id":141,"time":"2020-06-17 03:27:59.074381","s":48.7855297158},{"index":622,"x":348,"y":428,"id":3,"time":"2020-06-17 03:27:59.836655","s":49.5478036176},{"index":623,"x":791,"y":374,"id":4,"time":"2020-06-17 03:27:59.836655","s":49.5478036176},{"index":624,"x":789,"y":183,"id":96,"time":"2020-06-17 03:27:59.836655","s":49.5478036176},{"index":625,"x":255,"y":183,"id":99,"time":"2020-06-17 03:27:59.836655","s":49.5478036176},{"index":626,"x":635,"y":112,"id":108,"time":"2020-06-17 03:27:59.836655","s":49.5478036176},{"index":627,"x":348,"y":133,"id":114,"time":"2020-06-17 03:27:59.836655","s":49.5478036176},{"index":628,"x":545,"y":620,"id":123,"time":"2020-06-17 03:27:59.836655","s":49.5478036176},{"index":629,"x":512,"y":280,"id":124,"time":"2020-06-17 03:27:59.836655","s":49.5478036176},{"index":630,"x":103,"y":243,"id":137,"time":"2020-06-17 03:27:59.836655","s":49.5478036176},{"index":631,"x":254,"y":97,"id":143,"time":"2020-06-17 03:27:59.836655","s":49.5478036176},{"index":632,"x":374,"y":433,"id":3,"time":"2020-06-17 03:28:00.598929","s":50.3100775194},{"index":633,"x":626,"y":300,"id":4,"time":"2020-06-17 03:28:00.598929","s":50.3100775194},{"index":634,"x":254,"y":209,"id":99,"time":"2020-06-17 03:28:00.598929","s":50.3100775194},{"index":635,"x":665,"y":107,"id":108,"time":"2020-06-17 03:28:00.598929","s":50.3100775194},{"index":636,"x":348,"y":134,"id":114,"time":"2020-06-17 03:28:00.598929","s":50.3100775194},{"index":637,"x":539,"y":623,"id":123,"time":"2020-06-17 03:28:00.598929","s":50.3100775194},{"index":638,"x":492,"y":283,"id":124,"time":"2020-06-17 03:28:00.598929","s":50.3100775194},{"index":639,"x":107,"y":246,"id":137,"time":"2020-06-17 03:28:00.598929","s":50.3100775194},{"index":640,"x":256,"y":87,"id":143,"time":"2020-06-17 03:28:00.598929","s":50.3100775194},{"index":641,"x":371,"y":427,"id":3,"time":"2020-06-17 03:28:01.361202","s":51.0723514212},{"index":642,"x":584,"y":233,"id":4,"time":"2020-06-17 03:28:01.361202","s":51.0723514212},{"index":643,"x":274,"y":206,"id":99,"time":"2020-06-17 03:28:01.361202","s":51.0723514212},{"index":644,"x":663,"y":114,"id":108,"time":"2020-06-17 03:28:01.361202","s":51.0723514212},{"index":645,"x":350,"y":132,"id":114,"time":"2020-06-17 03:28:01.361202","s":51.0723514212},{"index":646,"x":522,"y":619,"id":123,"time":"2020-06-17 03:28:01.361202","s":51.0723514212},{"index":647,"x":487,"y":284,"id":124,"time":"2020-06-17 03:28:01.361202","s":51.0723514212},{"index":648,"x":95,"y":240,"id":137,"time":"2020-06-17 03:28:01.361202","s":51.0723514212},{"index":649,"x":254,"y":84,"id":143,"time":"2020-06-17 03:28:01.361202","s":51.0723514212},{"index":650,"x":371,"y":422,"id":3,"time":"2020-06-17 03:28:02.123476","s":51.834625323},{"index":651,"x":488,"y":153,"id":4,"time":"2020-06-17 03:28:02.123476","s":51.834625323},{"index":652,"x":280,"y":211,"id":99,"time":"2020-06-17 03:28:02.123476","s":51.834625323},{"index":653,"x":659,"y":112,"id":108,"time":"2020-06-17 03:28:02.123476","s":51.834625323},{"index":654,"x":521,"y":622,"id":123,"time":"2020-06-17 03:28:02.123476","s":51.834625323},{"index":655,"x":483,"y":281,"id":124,"time":"2020-06-17 03:28:02.123476","s":51.834625323},{"index":656,"x":87,"y":252,"id":137,"time":"2020-06-17 03:28:02.123476","s":51.834625323},{"index":657,"x":772,"y":145,"id":150,"time":"2020-06-17 03:28:02.123476","s":51.834625323},{"index":658,"x":367,"y":423,"id":3,"time":"2020-06-17 03:28:02.885750","s":52.5968992248},{"index":659,"x":433,"y":171,"id":4,"time":"2020-06-17 03:28:02.885750","s":52.5968992248},{"index":660,"x":290,"y":184,"id":99,"time":"2020-06-17 03:28:02.885750","s":52.5968992248},{"index":661,"x":659,"y":113,"id":108,"time":"2020-06-17 03:28:02.885750","s":52.5968992248},{"index":662,"x":350,"y":132,"id":114,"time":"2020-06-17 03:28:02.885750","s":52.5968992248},{"index":663,"x":517,"y":621,"id":123,"time":"2020-06-17 03:28:02.885750","s":52.5968992248},{"index":664,"x":490,"y":283,"id":124,"time":"2020-06-17 03:28:02.885750","s":52.5968992248},{"index":665,"x":95,"y":252,"id":137,"time":"2020-06-17 03:28:02.885750","s":52.5968992248},{"index":666,"x":253,"y":99,"id":143,"time":"2020-06-17 03:28:02.885750","s":52.5968992248},{"index":667,"x":765,"y":143,"id":150,"time":"2020-06-17 03:28:02.885750","s":52.5968992248},{"index":668,"x":359,"y":421,"id":3,"time":"2020-06-17 03:28:03.648024","s":53.3591731266},{"index":669,"x":404,"y":174,"id":4,"time":"2020-06-17 03:28:03.648024","s":53.3591731266},{"index":670,"x":284,"y":197,"id":99,"time":"2020-06-17 03:28:03.648024","s":53.3591731266},{"index":671,"x":657,"y":113,"id":108,"time":"2020-06-17 03:28:03.648024","s":53.3591731266},{"index":672,"x":490,"y":631,"id":123,"time":"2020-06-17 03:28:03.648024","s":53.3591731266},{"index":673,"x":495,"y":282,"id":124,"time":"2020-06-17 03:28:03.648024","s":53.3591731266},{"index":674,"x":257,"y":110,"id":143,"time":"2020-06-17 03:28:03.648024","s":53.3591731266},{"index":675,"x":1132,"y":577,"id":152,"time":"2020-06-17 03:28:03.648024","s":53.3591731266},{"index":676,"x":372,"y":429,"id":3,"time":"2020-06-17 03:28:04.410298","s":54.1214470284},{"index":677,"x":383,"y":157,"id":4,"time":"2020-06-17 03:28:04.410298","s":54.1214470284},{"index":678,"x":311,"y":201,"id":99,"time":"2020-06-17 03:28:04.410298","s":54.1214470284},{"index":679,"x":658,"y":114,"id":108,"time":"2020-06-17 03:28:04.410298","s":54.1214470284},{"index":680,"x":492,"y":286,"id":124,"time":"2020-06-17 03:28:04.410298","s":54.1214470284},{"index":681,"x":123,"y":250,"id":137,"time":"2020-06-17 03:28:04.410298","s":54.1214470284},{"index":682,"x":269,"y":139,"id":143,"time":"2020-06-17 03:28:04.410298","s":54.1214470284},{"index":683,"x":1024,"y":543,"id":152,"time":"2020-06-17 03:28:04.410298","s":54.1214470284},{"index":684,"x":237,"y":203,"id":155,"time":"2020-06-17 03:28:04.410298","s":54.1214470284},{"index":685,"x":368,"y":438,"id":3,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":686,"x":382,"y":110,"id":4,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":687,"x":365,"y":209,"id":99,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":688,"x":660,"y":114,"id":108,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":689,"x":490,"y":287,"id":124,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":690,"x":124,"y":251,"id":137,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":691,"x":280,"y":187,"id":143,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":692,"x":771,"y":141,"id":150,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":693,"x":1014,"y":527,"id":152,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":694,"x":621,"y":100,"id":157,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":695,"x":580,"y":664,"id":159,"time":"2020-06-17 03:28:05.172572","s":54.8837209302},{"index":696,"x":337,"y":469,"id":3,"time":"2020-06-17 03:28:05.934846","s":55.645994832},{"index":697,"x":385,"y":251,"id":99,"time":"2020-06-17 03:28:05.934846","s":55.645994832},{"index":698,"x":658,"y":114,"id":108,"time":"2020-06-17 03:28:05.934846","s":55.645994832},{"index":699,"x":490,"y":361,"id":124,"time":"2020-06-17 03:28:05.934846","s":55.645994832},{"index":700,"x":115,"y":252,"id":137,"time":"2020-06-17 03:28:05.934846","s":55.645994832},{"index":701,"x":300,"y":176,"id":143,"time":"2020-06-17 03:28:05.934846","s":55.645994832},{"index":702,"x":1011,"y":555,"id":152,"time":"2020-06-17 03:28:05.934846","s":55.645994832},{"index":703,"x":240,"y":167,"id":155,"time":"2020-06-17 03:28:05.934846","s":55.645994832},{"index":704,"x":620,"y":101,"id":157,"time":"2020-06-17 03:28:05.934846","s":55.645994832},{"index":705,"x":625,"y":582,"id":159,"time":"2020-06-17 03:28:05.934846","s":55.645994832},{"index":706,"x":283,"y":503,"id":3,"time":"2020-06-17 03:28:06.697120","s":56.4082687339},{"index":707,"x":640,"y":108,"id":108,"time":"2020-06-17 03:28:06.697120","s":56.4082687339},{"index":708,"x":417,"y":333,"id":124,"time":"2020-06-17 03:28:06.697120","s":56.4082687339},{"index":709,"x":105,"y":254,"id":137,"time":"2020-06-17 03:28:06.697120","s":56.4082687339},{"index":710,"x":314,"y":242,"id":143,"time":"2020-06-17 03:28:06.697120","s":56.4082687339},{"index":711,"x":763,"y":144,"id":150,"time":"2020-06-17 03:28:06.697120","s":56.4082687339},{"index":712,"x":968,"y":615,"id":152,"time":"2020-06-17 03:28:06.697120","s":56.4082687339},{"index":713,"x":724,"y":540,"id":159,"time":"2020-06-17 03:28:06.697120","s":56.4082687339},{"index":714,"x":437,"y":177,"id":162,"time":"2020-06-17 03:28:06.697120","s":56.4082687339},{"index":715,"x":253,"y":511,"id":3,"time":"2020-06-17 03:28:07.459394","s":57.1705426357},{"index":716,"x":472,"y":242,"id":99,"time":"2020-06-17 03:28:07.459394","s":57.1705426357},{"index":717,"x":634,"y":107,"id":108,"time":"2020-06-17 03:28:07.459394","s":57.1705426357},{"index":718,"x":370,"y":351,"id":124,"time":"2020-06-17 03:28:07.459394","s":57.1705426357},{"index":719,"x":107,"y":251,"id":137,"time":"2020-06-17 03:28:07.459394","s":57.1705426357},{"index":720,"x":316,"y":175,"id":143,"time":"2020-06-17 03:28:07.459394","s":57.1705426357},{"index":721,"x":764,"y":146,"id":150,"time":"2020-06-17 03:28:07.459394","s":57.1705426357},{"index":722,"x":239,"y":152,"id":155,"time":"2020-06-17 03:28:07.459394","s":57.1705426357},{"index":723,"x":722,"y":535,"id":159,"time":"2020-06-17 03:28:07.459394","s":57.1705426357},{"index":724,"x":712,"y":134,"id":165,"time":"2020-06-17 03:28:07.459394","s":57.1705426357},{"index":725,"x":272,"y":528,"id":3,"time":"2020-06-17 03:28:08.221668","s":57.9328165375},{"index":726,"x":537,"y":251,"id":99,"time":"2020-06-17 03:28:08.221668","s":57.9328165375},{"index":727,"x":639,"y":106,"id":108,"time":"2020-06-17 03:28:08.221668","s":57.9328165375},{"index":728,"x":361,"y":401,"id":124,"time":"2020-06-17 03:28:08.221668","s":57.9328165375},{"index":729,"x":315,"y":155,"id":143,"time":"2020-06-17 03:28:08.221668","s":57.9328165375},{"index":730,"x":764,"y":150,"id":150,"time":"2020-06-17 03:28:08.221668","s":57.9328165375},{"index":731,"x":762,"y":530,"id":159,"time":"2020-06-17 03:28:08.221668","s":57.9328165375},{"index":732,"x":288,"y":555,"id":3,"time":"2020-06-17 03:28:08.983941","s":58.6950904393},{"index":733,"x":584,"y":274,"id":99,"time":"2020-06-17 03:28:08.983941","s":58.6950904393},{"index":734,"x":609,"y":101,"id":108,"time":"2020-06-17 03:28:08.983941","s":58.6950904393},{"index":735,"x":365,"y":412,"id":124,"time":"2020-06-17 03:28:08.983941","s":58.6950904393},{"index":736,"x":108,"y":251,"id":137,"time":"2020-06-17 03:28:08.983941","s":58.6950904393},{"index":737,"x":314,"y":147,"id":143,"time":"2020-06-17 03:28:08.983941","s":58.6950904393},{"index":738,"x":764,"y":152,"id":150,"time":"2020-06-17 03:28:08.983941","s":58.6950904393},{"index":739,"x":780,"y":517,"id":159,"time":"2020-06-17 03:28:08.983941","s":58.6950904393},{"index":740,"x":694,"y":434,"id":168,"time":"2020-06-17 03:28:08.983941","s":58.6950904393},{"index":741,"x":709,"y":147,"id":169,"time":"2020-06-17 03:28:08.983941","s":58.6950904393}]}
\ No newline at end of file

+{"schema":{"fields":[{"name":"index","type":"string"},{"name":"total","type":"string"},{"name":"now","type":"string"},{"name":"time","type":"string"},{"name":"s","type":"string"}],"primaryKey":["index"],"pandas_version":"0.20.0"},"data":[]}
\ No newline at end of file

+{"schema":{"fields":[{"name":"index","type":"string"},{"name":"x","type":"string"},{"name":"y","type":"string"},{"name":"id","type":"string"},{"name":"time","type":"string"},{"name":"s","type":"string"}],"primaryKey":["index"],"pandas_version":"0.20.0"},"data":[]}
\ No newline at end of file

+Keras==2.3.1
+tensorflow-gpu==1.13.1
+numpy==1.15.0
+opencv-python==4.2.0.34
+scikit-learn==0.23.1
+scipy==1.4.1
+Pillow
+torch==1.4.0
+torchvision==0.5.0

+Intput_video
+

+import cv2
+import os
+import moviepy.editor as mp
+
+video = "00046"
+#input_imgs
+im_dir = '/home/cai/Desktop/yolo_dataset/t1_video/t1_video_'+video+'/'
+#im_dir =
+#output_video
+video_dir = '/home/cai/Desktop/yolo_dataset/t1_video/test_video/det_t1_video_'+video+'_test_q.avi'
+#fps
+fps = 50
+#num_of_imgs
+num = 310
+#img_size
+img_size = (1920,1080)
+fourcc = cv2.VideoWriter_fourcc('M','J','P','G')
+#opencv3
+
+videoWriter = cv2.VideoWriter(video_dir, fourcc, fps, img_size)
+for i in range(0,num):
+    #im_name = os.path.join(im_dir,'frame-' + str(i) + '.png')
+    im_name = os.path.join(im_dir,'t1_video_'+video+'_' + "%05d" % i + '.jpg')
+    frame = cv2.imread(im_name)
+    #frame = cv2.resize(frame, (480, 320))
+    #frame = cv2.resize(frame,(520,320), interpolation=cv2.INTER_CUBIC)
+    videoWriter.write(frame)
+    print (im_name)
+videoWriter.release()
+print('finish')

+import argparse
+import tensorflow as tf
+import tensorflow.contrib.slim as slim
+
+
+def _batch_norm_fn(x, scope=None):
+    if scope is None:
+        scope = tf.get_variable_scope().name + "/bn"
+    return slim.batch_norm(x, scope=scope)
+
+
+def create_link(
+        incoming, network_builder, scope, nonlinearity=tf.nn.elu,
+        weights_initializer=tf.truncated_normal_initializer(stddev=1e-3),
+        regularizer=None, is_first=False, summarize_activations=True):
+    if is_first:
+        network = incoming
+    else:
+        network = _batch_norm_fn(incoming, scope=scope + "/bn")
+        network = nonlinearity(network)
+        if summarize_activations:
+            tf.summary.histogram(scope+"/activations", network)
+
+    pre_block_network = network
+    post_block_network = network_builder(pre_block_network, scope)
+
+    incoming_dim = pre_block_network.get_shape().as_list()[-1]
+    outgoing_dim = post_block_network.get_shape().as_list()[-1]
+    if incoming_dim != outgoing_dim:
+        assert outgoing_dim == 2 * incoming_dim, \
+            "%d != %d" % (outgoing_dim, 2 * incoming)
+        projection = slim.conv2d(
+            incoming, outgoing_dim, 1, 2, padding="SAME", activation_fn=None,
+            scope=scope+"/projection", weights_initializer=weights_initializer,
+            biases_initializer=None, weights_regularizer=regularizer)
+        network = projection + post_block_network
+    else:
+        network = incoming + post_block_network
+    return network
+
+
+def create_inner_block(
+        incoming, scope, nonlinearity=tf.nn.elu,
+        weights_initializer=tf.truncated_normal_initializer(1e-3),
+        bias_initializer=tf.zeros_initializer(), regularizer=None,
+        increase_dim=False, summarize_activations=True):
+    n = incoming.get_shape().as_list()[-1]
+    stride = 1
+    if increase_dim:
+        n *= 2
+        stride = 2
+
+    incoming = slim.conv2d(
+        incoming, n, [3, 3], stride, activation_fn=nonlinearity, padding="SAME",
+        normalizer_fn=_batch_norm_fn, weights_initializer=weights_initializer,
+        biases_initializer=bias_initializer, weights_regularizer=regularizer,
+        scope=scope + "/1")
+    if summarize_activations:
+        tf.summary.histogram(incoming.name + "/activations", incoming)
+
+    incoming = slim.dropout(incoming, keep_prob=0.6)
+
+    incoming = slim.conv2d(
+        incoming, n, [3, 3], 1, activation_fn=None, padding="SAME",
+        normalizer_fn=None, weights_initializer=weights_initializer,
+        biases_initializer=bias_initializer, weights_regularizer=regularizer,
+        scope=scope + "/2")
+    return incoming
+
+
+def residual_block(incoming, scope, nonlinearity=tf.nn.elu,
+                   weights_initializer=tf.truncated_normal_initializer(1e3),
+                   bias_initializer=tf.zeros_initializer(), regularizer=None,
+                   increase_dim=False, is_first=False,
+                   summarize_activations=True):
+
+    def network_builder(x, s):
+        return create_inner_block(
+            x, s, nonlinearity, weights_initializer, bias_initializer,
+            regularizer, increase_dim, summarize_activations)
+
+    return create_link(
+        incoming, network_builder, scope, nonlinearity, weights_initializer,
+        regularizer, is_first, summarize_activations)
+
+
+def _create_network(incoming, reuse=None, weight_decay=1e-8):
+    nonlinearity = tf.nn.elu
+    conv_weight_init = tf.truncated_normal_initializer(stddev=1e-3)
+    conv_bias_init = tf.zeros_initializer()
+    conv_regularizer = slim.l2_regularizer(weight_decay)
+    fc_weight_init = tf.truncated_normal_initializer(stddev=1e-3)
+    fc_bias_init = tf.zeros_initializer()
+    fc_regularizer = slim.l2_regularizer(weight_decay)
+
+    def batch_norm_fn(x):
+        return slim.batch_norm(x, scope=tf.get_variable_scope().name + "/bn")
+
+    network = incoming
+    network = slim.conv2d(
+        network, 32, [3, 3], stride=1, activation_fn=nonlinearity,
+        padding="SAME", normalizer_fn=batch_norm_fn, scope="conv1_1",
+        weights_initializer=conv_weight_init, biases_initializer=conv_bias_init,
+        weights_regularizer=conv_regularizer)
+    network = slim.conv2d(
+        network, 32, [3, 3], stride=1, activation_fn=nonlinearity,
+        padding="SAME", normalizer_fn=batch_norm_fn, scope="conv1_2",
+        weights_initializer=conv_weight_init, biases_initializer=conv_bias_init,
+        weights_regularizer=conv_regularizer)
+
+    # NOTE(nwojke): This is missing a padding="SAME" to match the CNN
+    # architecture in Table 1 of the paper. Information on how this affects
+    # performance on MOT 16 training sequences can be found in
+    # issue 10 https://github.com/nwojke/deep_sort/issues/10
+    network = slim.max_pool2d(network, [3, 3], [2, 2], scope="pool1")
+
+    network = residual_block(
+        network, "conv2_1", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False, is_first=True)
+    network = residual_block(
+        network, "conv2_3", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False)
+
+    network = residual_block(
+        network, "conv3_1", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=True)
+    network = residual_block(
+        network, "conv3_3", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False)
+
+    network = residual_block(
+        network, "conv4_1", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=True)
+    network = residual_block(
+        network, "conv4_3", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False)
+
+    feature_dim = network.get_shape().as_list()[-1]
+    network = slim.flatten(network)
+
+    network = slim.dropout(network, keep_prob=0.6)
+    network = slim.fully_connected(
+        network, feature_dim, activation_fn=nonlinearity,
+        normalizer_fn=batch_norm_fn, weights_regularizer=fc_regularizer,
+        scope="fc1", weights_initializer=fc_weight_init,
+        biases_initializer=fc_bias_init)
+
+    features = network
+
+    # Features in rows, normalize axis 1.
+    features = slim.batch_norm(features, scope="ball", reuse=reuse)
+    feature_norm = tf.sqrt(
+        tf.constant(1e-8, tf.float32) +
+        tf.reduce_sum(tf.square(features), [1], keepdims=True))
+    features = features / feature_norm
+    return features, None
+
+
+def _network_factory(weight_decay=1e-8):
+
+    def factory_fn(image, reuse):
+            with slim.arg_scope([slim.batch_norm, slim.dropout],
+                                is_training=False):
+                with slim.arg_scope([slim.conv2d, slim.fully_connected,
+                                     slim.batch_norm, slim.layer_norm],
+                                    reuse=reuse):
+                    features, logits = _create_network(
+                        image, reuse=reuse, weight_decay=weight_decay)
+                    return features, logits
+
+    return factory_fn
+
+
+def _preprocess(image):
+    image = image[:, :, ::-1]  # BGR to RGB
+    return image
+
+
+def parse_args():
+    """Parse command line arguments.
+    """
+    parser = argparse.ArgumentParser(description="Freeze old model")
+    parser.add_argument(
+        "--checkpoint_in",
+        default="resources/networks/mars-small128.ckpt-68577",
+        help="Path to checkpoint file")
+    parser.add_argument(
+        "--graphdef_out",
+        default="resources/networks/mars-small128.pb")
+    return parser.parse_args()
+
+
+def main():
+    args = parse_args()
+
+    with tf.Session(graph=tf.Graph()) as session:
+        input_var = tf.placeholder(
+            tf.uint8, (None, 128, 64, 3), name="images")
+        image_var = tf.map_fn(
+            lambda x: _preprocess(x), tf.cast(input_var, tf.float32),
+            back_prop=False)
+
+        factory_fn = _network_factory()
+        features, _ = factory_fn(image_var, reuse=None)
+        features = tf.identity(features, name="features")
+
+        saver = tf.train.Saver(slim.get# vim: expandtab:ts=4:sw=4
+import argparse
+import tensorflow as tf
+import tensorflow.contrib.slim as slim
+
+
+def _batch_norm_fn(x, scope=None):
+    if scope is None:
+        scope = tf.get_variable_scope().name + "/bn"
+    return slim.batch_norm(x, scope=scope)
+
+
+def create_link(
+        incoming, network_builder, scope, nonlinearity=tf.nn.elu,
+        weights_initializer=tf.truncated_normal_initializer(stddev=1e-3),
+        regularizer=None, is_first=False, summarize_activations=True):
+    if is_first:
+        network = incoming
+    else:
+        network = _batch_norm_fn(incoming, scope=scope + "/bn")
+        network = nonlinearity(network)
+        if summarize_activations:
+            tf.summary.histogram(scope+"/activations", network)
+
+    pre_block_network = network
+    post_block_network = network_builder(pre_block_network, scope)
+
+    incoming_dim = pre_block_network.get_shape().as_list()[-1]
+    outgoing_dim = post_block_network.get_shape().as_list()[-1]
+    if incoming_dim != outgoing_dim:
+        assert outgoing_dim == 2 * incoming_dim, \
+            "%d != %d" % (outgoing_dim, 2 * incoming)
+        projection = slim.conv2d(
+            incoming, outgoing_dim, 1, 2, padding="SAME", activation_fn=None,
+            scope=scope+"/projection", weights_initializer=weights_initializer,
+            biases_initializer=None, weights_regularizer=regularizer)
+        network = projection + post_block_network
+    else:
+        network = incoming + post_block_network
+    return network
+
+
+def create_inner_block(
+        incoming, scope, nonlinearity=tf.nn.elu,
+        weights_initializer=tf.truncated_normal_initializer(1e-3),
+        bias_initializer=tf.zeros_initializer(), regularizer=None,
+        increase_dim=False, summarize_activations=True):
+    n = incoming.get_shape().as_list()[-1]
+    stride = 1
+    if increase_dim:
+        n *= 2
+        stride = 2
+
+    incoming = slim.conv2d(
+        incoming, n, [3, 3], stride, activation_fn=nonlinearity, padding="SAME",
+        normalizer_fn=_batch_norm_fn, weights_initializer=weights_initializer,
+        biases_initializer=bias_initializer, weights_regularizer=regularizer,
+        scope=scope + "/1")
+    if summarize_activations:
+        tf.summary.histogram(incoming.name + "/activations", incoming)
+
+    incoming = slim.dropout(incoming, keep_prob=0.6)
+
+    incoming = slim.conv2d(
+        incoming, n, [3, 3], 1, activation_fn=None, padding="SAME",
+        normalizer_fn=None, weights_initializer=weights_initializer,
+        biases_initializer=bias_initializer, weights_regularizer=regularizer,
+        scope=scope + "/2")
+    return incoming
+
+
+def residual_block(incoming, scope, nonlinearity=tf.nn.elu,
+                   weights_initializer=tf.truncated_normal_initializer(1e3),
+                   bias_initializer=tf.zeros_initializer(), regularizer=None,
+                   increase_dim=False, is_first=False,
+                   summarize_activations=True):
+
+    def network_builder(x, s):
+        return create_inner_block(
+            x, s, nonlinearity, weights_initializer, bias_initializer,
+            regularizer, increase_dim, summarize_activations)
+
+    return create_link(
+        incoming, network_builder, scope, nonlinearity, weights_initializer,
+        regularizer, is_first, summarize_activations)
+
+
+def _create_network(incoming, reuse=None, weight_decay=1e-8):
+    nonlinearity = tf.nn.elu
+    conv_weight_init = tf.truncated_normal_initializer(stddev=1e-3)
+    conv_bias_init = tf.zeros_initializer()
+    conv_regularizer = slim.l2_regularizer(weight_decay)
+    fc_weight_init = tf.truncated_normal_initializer(stddev=1e-3)
+    fc_bias_init = tf.zeros_initializer()
+    fc_regularizer = slim.l2_regularizer(weight_decay)
+
+    def batch_norm_fn(x):
+        return slim.batch_norm(x, scope=tf.get_variable_scope().name + "/bn")
+
+    network = incoming
+    network = slim.conv2d(
+        network, 32, [3, 3], stride=1, activation_fn=nonlinearity,
+        padding="SAME", normalizer_fn=batch_norm_fn, scope="conv1_1",
+        weights_initializer=conv_weight_init, biases_initializer=conv_bias_init,
+        weights_regularizer=conv_regularizer)
+    network = slim.conv2d(
+        network, 32, [3, 3], stride=1, activation_fn=nonlinearity,
+        padding="SAME", normalizer_fn=batch_norm_fn, scope="conv1_2",
+        weights_initializer=conv_weight_init, biases_initializer=conv_bias_init,
+        weights_regularizer=conv_regularizer)
+
+    # NOTE(nwojke): This is missing a padding="SAME" to match the CNN
+    # architecture in Table 1 of the paper. Information on how this affects
+    # performance on MOT 16 training sequences can be found in
+    # issue 10 https://github.com/nwojke/deep_sort/issues/10
+    network = slim.max_pool2d(network, [3, 3], [2, 2], scope="pool1")
+
+    network = residual_block(
+        network, "conv2_1", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False, is_first=True)
+    network = residual_block(
+        network, "conv2_3", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False)
+
+    network = residual_block(
+        network, "conv3_1", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=True)
+    network = residual_block(
+        network, "conv3_3", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False)
+
+    network = residual_block(
+        network, "conv4_1", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=True)
+    network = residual_block(
+        network, "conv4_3", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False)
+
+    feature_dim = network.get_shape().as_list()[-1]
+    network = slim.flatten(network)
+
+    network = slim.dropout(network, keep_prob=0.6)
+    network = slim.fully_connected(
+        network, feature_dim, activation_fn=nonlinearity,
+        normalizer_fn=batch_norm_fn, weights_regularizer=fc_regularizer,
+        scope="fc1", weights_initializer=fc_weight_init,
+        biases_initializer=fc_bias_init)
+
+    features = network
+
+    # Features in rows, normalize axis 1.
+    features = slim.batch_norm(features, scope="ball", reuse=reuse)
+    feature_norm = tf.sqrt(
+        tf.constant(1e-8, tf.float32) +
+        tf.reduce_sum(tf.square(features), [1], keepdims=True))
+    features = features / feature_norm
+    return features, None
+
+
+def _network_factory(weight_decay=1e-8):
+
+    def factory_fn(image, reuse):
+            with slim.arg_scope([slim.batch_norm, slim.dropout],
+                                is_training=False):
+                with slim.arg_scope([slim.conv2d, slim.fully_connected,
+                                     slim.batch_norm, slim.layer_norm],
+                                    reuse=reuse):
+                    features, logits = _create_network(
+                        image, reuse=reuse, weight_decay=weight_decay)
+                    return features, logits
+
+    return factory_fn
+
+
+def _preprocess(image):
+    image = image[:, :, ::-1]  # BGR to RGB
+    return image
+
+
+def parse_args():
+    """Parse command line arguments.
+    """
+    parser = argparse.ArgumentParser(description="Freeze old model")
+    parser.add_argument(
+        "--checkpoint_in",# vim: expandtab:ts=4:sw=4
+import argparse
+import tensorflow as tf
+import tensorflow.contrib.slim as slim
+
+
+def _batch_norm_fn(x, scope=None):
+    if scope is None:
+        scope = tf.get_variable_scope().name + "/bn"
+    return slim.batch_norm(x, scope=scope)
+
+
+def create_link(
+        incoming, network_builder, scope, nonlinearity=tf.nn.elu,
+        weights_initializer=tf.truncated_normal_initializer(stddev=1e-3),
+        regularizer=None, is_first=False, summarize_activations=True):
+    if is_first:
+        network = incoming
+    else:
+        network = _batch_norm_fn(incoming, scope=scope + "/bn")
+        network = nonlinearity(network)
+        if summarize_activations:
+            tf.summary.histogram(scope+"/activations", network)
+
+    pre_block_network = network
+    post_block_network = network_builder(pre_block_network, scope)
+
+    incoming_dim = pre_block_network.get_shape().as_list()[-1]
+    outgoing_dim = post_block_network.get_shape().as_list()[-1]
+    if incoming_dim != outgoing_dim:
+        assert outgoing_dim == 2 * incoming_dim, \
+            "%d != %d" % (outgoing_dim, 2 * incoming)
+        projection = slim.conv2d(
+            incoming, outgoing_dim, 1, 2, padding="SAME", activation_fn=None,
+            scope=scope+"/projection", weights_initializer=weights_initializer,
+            biases_initializer=None, weights_regularizer=regularizer)
+        network = projection + post_block_network
+    else:
+        network = incoming + post_block_network
+    return network
+
+
+def create_inner_block(
+        incoming, scope, nonlinearity=tf.nn.elu,
+        weights_initializer=tf.truncated_normal_initializer(1e-3),
+        bias_initializer=tf.zeros_initializer(), regularizer=None,
+        increase_dim=False, summarize_activations=True):
+    n = incoming.get_shape().as_list()[-1]
+    stride = 1
+    if increase_dim:
+        n *= 2
+        stride = 2
+
+    incoming = slim.conv2d(
+        incoming, n, [3, 3], stride, activation_fn=nonlinearity, padding="SAME",
+        normalizer_fn=_batch_norm_fn, weights_initializer=weights_initializer,
+        biases_initializer=bias_initializer, weights_regularizer=regularizer,
+        scope=scope + "/1")
+    if summarize_activations:
+        tf.summary.histogram(incoming.name + "/activations", incoming)
+
+    incoming = slim.dropout(incoming, keep_prob=0.6)
+
+    incoming = slim.conv2d(
+        incoming, n, [3, 3], 1, activation_fn=None, padding="SAME",
+        normalizer_fn=None, weights_initializer=weights_initializer,
+        biases_initializer=bias_initializer, weights_regularizer=regularizer,
+        scope=scope + "/2")
+    return incoming
+
+
+def residual_block(incoming, scope, nonlinearity=tf.nn.elu,
+                   weights_initializer=tf.truncated_normal_initializer(1e3),
+                   bias_initializer=tf.zeros_initializer(), regularizer=None,
+                   increase_dim=False, is_first=False,
+                   summarize_activations=True):
+
+    def network_builder(x, s):
+        return create_inner_block(
+            x, s, nonlinearity, weights_initializer, bias_initializer,
+            regularizer, increase_dim, summarize_activations)
+
+    return create_link(
+        incoming, network_builder, scope, nonlinearity, weights_initializer,
+        regularizer, is_first, summarize_activations)
+
+
+def _create_network(incoming, reuse=None, weight_decay=1e-8):
+    nonlinearity = tf.nn.elu
+    conv_weight_init = tf.truncated_normal_initializer(stddev=1e-3)
+    conv_bias_init = tf.zeros_initializer()
+    conv_regularizer = slim.l2_regularizer(weight_decay)
+    fc_weight_init = tf.truncated_normal_initializer(stddev=1e-3)
+    fc_bias_init = tf.zeros_initializer()
+    fc_regularizer = slim.l2_regularizer(weight_decay)
+
+    def batch_norm_fn(x):
+        return slim.batch_norm(x, scope=tf.get_variable_scope().name + "/bn")
+
+    network = incoming
+    network = slim.conv2d(
+        network, 32, [3, 3], stride=1, activation_fn=nonlinearity,
+        padding="SAME", normalizer_fn=batch_norm_fn, scope="conv1_1",
+        weights_initializer=conv_weight_init, biases_initializer=conv_bias_init,
+        weights_regularizer=conv_regularizer)
+    network = slim.conv2d(
+        network, 32, [3, 3], stride=1, activation_fn=nonlinearity,
+        padding="SAME", normalizer_fn=batch_norm_fn, scope="conv1_2",
+        weights_initializer=conv_weight_init, biases_initializer=conv_bias_init,
+        weights_regularizer=conv_regularizer)
+
+    # NOTE(nwojke): This is missing a padding="SAME" to match the CNN
+    # architecture in Table 1 of the paper. Information on how this affects
+    # performance on MOT 16 training sequences can be found in
+    # issue 10 https://github.com/nwojke/deep_sort/issues/10
+    network = slim.max_pool2d(network, [3, 3], [2, 2], scope="pool1")
+
+    network = residual_block(
+        network, "conv2_1", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False, is_first=True)
+    network = residual_block(
+        network, "conv2_3", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False)
+
+    network = residual_block(
+        network, "conv3_1", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=True)
+    network = residual_block(
+        network, "conv3_3", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False)
+
+    network = residual_block(
+        network, "conv4_1", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=True)
+    network = residual_block(
+        network, "conv4_3", nonlinearity, conv_weight_init, conv_bias_init,
+        conv_regularizer, increase_dim=False)
+
+    feature_dim = network.get_shape().as_list()[-1]
+    network = slim.flatten(network)
+
+    network = slim.dropout(network, keep_prob=0.6)
+    network = slim.fully_connected(
+        network, feature_dim, activation_fn=nonlinearity,
+        normalizer_fn=batch_norm_fn, weights_regularizer=fc_regularizer,
+        scope="fc1", weights_initializer=fc_weight_init,
+        biases_initializer=fc_bias_init)
+
+    features = network
+
+    # Features in rows, normalize axis 1.
+    features = slim.batch_norm(features, scope="ball", reuse=reuse)
+    feature_norm = tf.sqrt(
+        tf.constant(1e-8, tf.float32) +
+        tf.reduce_sum(tf.square(features), [1], keepdims=True))
+    features = features / feature_norm
+    return features, None
+
+
+def _network_factory(weight_decay=1e-8):
+
+    def factory_fn(image, reuse):
+            with slim.arg_scope([slim.batch_norm, slim.dropout],
+                                is_training=False):
+                with slim.arg_scope([slim.conv2d, slim.fully_connected,
+                                     slim.batch_norm, slim.layer_norm],
+                                    reuse=reuse):
+                    features, logits = _create_network(
+                        image, reuse=reuse, weight_decay=weight_decay)
+                    return features, logits
+
+    return factory_fn
+
+
+def _preprocess(image):
+    image = image[:, :, ::-1]  # BGR to RGB
+    return image
+
+
+def parse_args():
+    """Parse command line arguments.
+    """
+    parser = argparse.ArgumentParser(description="Freeze old model")
+    parser.add_argument(
+        "--checkpoint_in",
+        default="resources/networks/mars-small128.ckpt-68577",
+        help="Path to checkpoint file")
+    parser.add_argument(
+        "--graphdef_out",
+        default="resources/networks/mars-small128.pb")
+    return parser.parse_args()
+
+
+def main():
+    args = parse_args()
+
+    with tf.Session(graph=tf.Graph()) as session:
+        input_var = tf.placeholder(
+            tf.uint8, (None, 128, 64, 3), name="images")
+        image_var = tf.map_fn(
+            lambda x: _preprocess(x), tf.cast(input_var, tf.float32),
+            back_prop=False)
+
+        factory_fn = _network_factory()
+        features, _ = factory_fn(image_var, reuse=None)
+        features = tf.identity(features, name="features")
+
+        saver = tf.train.Saver(slim.get_variables_to_restore())
+        saver.restore(session, args.checkpoint_in)
+
+        output_graph_def = tf.graph_util.convert_variables_to_constants(
+            session, tf.get_default_graph().as_graph_def(),
+            [features.name.split(":")[0]])
+        with tf.gfile.GFile(args.graphdef_out, "wb") as file_handle:
+            file_handle.write(output_graph_def.SerializeToString())
+
+
+if __name__ == "__main__":
+        default="resources/networks/mars-small128.ckpt-68577",
+        help="Path to checkpoint file")
+    parser.add_argument(
+        "--graphdef_out",
+        default="resources/networks/mars-small128.pb")
+    return parser.parse_args()
+
+
+def main():
+    args = parse_args()
+
+    with tf.Session(graph=tf.Graph()) as session:
+        input_var = tf.placeholder(
+            tf.uint8, (None, 128, 64, 3), name="images")
+        image_var = tf.map_fn(
+            lambda x: _preprocess(x), tf.cast(input_var, tf.float32),
+            back_prop=False)
+
+        factory_fn = _network_factory()
+        features, _ = factory_fn(image_var, reuse=None)
+        features = tf.identity(features, name="features")
+
+        saver = tf.train.Saver(slim.get_variables_to_restore())
+        saver.restore(session, args.checkpoint_in)
+
+        output_graph_def = tf.graph_util.convert_variables_to_constants(
+            session, tf.get_default_graph().as_graph_def(),
+            [features.name.split(":")[0]])
+        with tf.gfile.GFile(args.graphdef_out, "wb") as file_handle:
+            file_handle.write(output_graph_def.SerializeToString())
+
+
+if __name__ == "__main__":_variables_to_restore())
+        saver.restore(session, args.checkpoint_in)
+
+        output_graph_def = tf.graph_util.convert_variables_to_constants(
+            session, tf.get_default_graph().as_graph_def(),
+            [features.name.split(":")[0]])
+        with tf.gfile.GFile(args.graphdef_out, "wb") as file_handle:
+            file_handle.write(output_graph_def.SerializeToString())
+
+
+if __name__ == "__main__":
+    main()

+import os
+import errno
+import argparse
+import numpy as np
+import cv2
+import tensorflow as tf
+
+
+def _run_in_batches(f, data_dict, out, batch_size):
+    data_len = len(out)
+    num_batches = int(data_len / batch_size)
+
+    s, e = 0, 0
+    for i in range(num_batches):
+        s, e = i * batch_size, (i + 1) * batch_size
+        batch_data_dict = {k: v[s:e] for k, v in data_dict.items()}
+        out[s:e] = f(batch_data_dict)
+    if e < len(out):
+        batch_data_dict = {k: v[e:] for k, v in data_dict.items()}
+        out[e:] = f(batch_data_dict)
+
+
+def extract_image_patch(image, bbox, patch_shape):
+    """Extract image patch from bounding box.
+    Parameters
+    ----------
+    image : ndarray
+        The full image.
+    bbox : array_like
+        The bounding box in format (x, y, width, height).
+    patch_shape : Optional[array_like]
+        This parameter can be used to enforce a desired patch shape
+        (height, width). First, the `bbox` is adapted to the aspect ratio
+        of the patch shape, then it is clipped at the image boundaries.
+        If None, the shape is computed from :arg:`bbox`.
+    Returns
+    -------
+    ndarray | NoneType
+        An image patch showing the :arg:`bbox`, optionally reshaped to
+        :arg:`patch_shape`.
+        Returns None if the bounding box is empty or fully outside of the image
+        boundaries.
+    """
+    bbox = np.array(bbox)
+    if patch_shape is not None:
+        # correct aspect ratio to patch shape
+        target_aspect = float(patch_shape[1]) / patch_shape[0]
+        new_width = target_aspect * bbox[3]
+        bbox[0] -= (new_width - bbox[2]) / 2
+        bbox[2] = new_width
+
+    # convert to top left, bottom right
+    bbox[2:] += bbox[:2]
+    bbox = bbox.astype(np.int)
+
+    # clip at image boundaries
+    bbox[:2] = np.maximum(0, bbox[:2])
+    bbox[2:] = np.minimum(np.asarray(image.shape[:2][::-1]) - 1, bbox[2:])
+    if np.any(bbox[:2] >= bbox[2:]):
+        return None
+    sx, sy, ex, ey = bbox
+    image = image[sy:ey, sx:ex]
+    image = cv2.resize(image, tuple(patch_shape[::-1]))
+    return image
+
+
+class ImageEncoder(object):
+
+    def __init__(self, checkpoint_filename, input_name="images",
+                 output_name="features"):
+        self.session = tf.Session()
+        with tf.gfile.GFile(checkpoint_filename, "rb") as file_handle:
+            graph_def = tf.GraphDef()
+            graph_def.ParseFromString(file_handle.read())
+        tf.import_graph_def(graph_def, name="net")
+        self.input_var = tf.get_default_graph().get_tensor_by_name(
+            "net/%s:0" % input_name)
+        self.output_var = tf.get_default_graph().get_tensor_by_name(
+            "net/%s:0" % output_name)
+
+        assert len(self.output_var.get_shape()) == 2
+        assert len(self.input_var.get_shape()) == 4
+        self.feature_dim = self.output_var.get_shape().as_list()[-1]
+        self.image_shape = self.input_var.get_shape().as_list()[1:]
+
+    def __call__(self, data_x, batch_size=32):
+        out = np.zeros((len(data_x), self.feature_dim), np.float32)
+        _run_in_batches(
+            lambda x: self.session.run(self.output_var, feed_dict=x),
+            {self.input_var: data_x}, out, batch_size)
+        return out
+
+
+def create_box_encoder(model_filename, input_name="images",
+                       output_name="features", batch_size=32):
+    image_encoder = ImageEncoder(model_filename, input_name, output_name)
+    image_shape = image_encoder.image_shape
+
+    def encoder(image, boxes):
+        image_patches = []
+        for box in boxes:
+            patch = extract_image_patch(image, box, image_shape[:2])
+            if patch is None:
+                print("WARNING: Failed to extract image patch: %s." % str(box))
+                patch = np.random.uniform(
+                    0., 255., image_shape).astype(np.uint8)
+            image_patches.append(patch)
+        image_patches = np.asarray(image_patches)
+        return image_encoder(image_patches, batch_size)
+
+    return encoder
+
+
+def generate_detections(encoder, mot_dir, output_dir, detection_dir=None):
+    """Generate detections with features.
+    Parameters
+    ----------
+    encoder : Callable[image, ndarray] -> ndarray
+        The encoder function takes as input a BGR color image and a matrix of
+        bounding boxes in format `(x, y, w, h)` and returns a matrix of
+        corresponding feature vectors.
+    mot_dir : str
+        Path to the MOTChallenge directory (can be either train or test).
+    output_dir
+        Path to the output directory. Will be created if it does not exist.
+    detection_dir
+        Path to custom detections. The directory structure should be the default
+        MOTChallenge structure: `[sequence]/det/det.txt`. If None, uses the
+        standard MOTChallenge detections.
+    """
+    if detection_dir is None:
+        detection_dir = mot_dir
+    try:
+        os.makedirs(output_dir)
+    except OSError as exception:
+        if exception.errno == errno.EEXIST and os.path.isdir(output_dir):
+            pass
+        else:
+            raise ValueError(
+                "Failed to created output directory '%s'" % output_dir)
+
+    for sequence in os.listdir(mot_dir):
+        print("Processing %s" % sequence)
+        sequence_dir = os.path.join(mot_dir, sequence)
+
+        image_dir = os.path.join(sequence_dir, "img1")
+        image_filenames = {
+            int(os.path.splitext(f)[0]): os.path.join(image_dir, f)
+            for f in os.listdir(image_dir)}
+
+        detection_file = os.path.join(
+            detection_dir, sequence, "det/det.txt")
+        detections_in = np.loadtxt(detection_file, delimiter=',')
+        detections_out = []
+
+        frame_indices = detections_in[:, 0].astype(np.int)
+        min_frame_idx = frame_indices.astype(np.int).min()
+        max_frame_idx = frame_indices.astype(np.int).max()
+        for frame_idx in range(min_frame_idx, max_frame_idx + 1):
+            print("Frame %05d/%05d" % (frame_idx, max_frame_idx))
+            mask = frame_indices == frame_idx
+            rows = detections_in[mask]
+
+            if frame_idx not in image_filenames:
+                print("WARNING could not find image for frame %d" % frame_idx)
+                continue
+            bgr_image = cv2.imread(
+                image_filenames[frame_idx], cv2.IMREAD_COLOR)
+            features = encoder(bgr_image, rows[:, 2:6].copy())
+            detections_out += [np.r_[(row, feature)] for row, feature
+                               in zip(rows, features)]
+
+        output_filename = os.path.join(output_dir, "%s.npy" % sequence)
+        np.save(
+            output_filename, np.asarray(detections_out), allow_pickle=False)
+
+
+def parse_args():
+    """Parse command line arguments.
+    """
+    parser = argparse.ArgumentParser(description="Re-ID feature extractor")
+    parser.add_argument(
+        "--model",
+        default="resources/networks/mars-small128.pb",
+        help="Path to freezed inference graph protobuf.")
+    parser.add_argument(
+        "--mot_dir", help="Path to MOTChallenge directory (train or test)",
+        required=True)
+    parser.add_argument(
+        "--detection_dir", help="Path to custom detections. Defaults to "
+        "standard MOT detections Directory structure should be the default "
+        "MOTChallenge structure: [sequence]/det/det.txt", default=None)
+    parser.add_argument(
+        "--output_dir", help="Output directory. Will be created if it does not"
+        " exist.", default="detections")
+    return parser.parse_args()
+
+
+def main():
+    args = parse_args()
+    encoder = create_box_encoder(args.model, batch_size=32)
+    generate_detections(encoder, args.mot_dir, args.output_dir,
+                        args.detection_dir)
+
+
+if __name__ == "__main__":
+    main()

+import cv2
+
+image_folder = './mask_face'
+video_name = './cut_test.m4v'
+
+vc = cv2.VideoCapture(video_name)
+c = 1
+if vc.isOpened():
+    rval,frame=vc.read()
+else:
+    rval=False
+while rval:
+    rval,frame=vc.read()
+    cv2.imwrite('./mask_face/IMG_'+str(c)+'.jpg',frame)
+    c=c+1
+    cv2.waitKey(1)
+vc.release()

+import colorsys
+
+import numpy as np
+from keras import backend as K
+from keras.models import load_model
+
+from yolo4.model import yolo_eval, Mish
+from yolo4.utils import letterbox_image
+import os
+from keras.utils import multi_gpu_model
+
+class YOLO(object):
+    def __init__(self):
+        self.model_path = os.getcwd() + '/../deep_sort_yolov4/model_data/yolo4_weight.h5'
+        self.anchors_path = os.getcwd() + '/../deep_sort_yolov4/model_data/yolo_anchors.txt'
+        self.classes_path = os.getcwd() + '/../deep_sort_yolov4/model_data/coco_classes.txt'
+        self.gpu_num = 1
+        self.score = 0.5
+        self.iou = 0.5
+        self.class_names = self._get_class()
+        self.anchors = self._get_anchors()
+        self.sess = K.get_session()
+        self.model_image_size = (608, 608)  # fixed size or (None, None)
+        self.is_fixed_size = self.model_image_size != (None, None)
+        self.boxes, self.scores, self.classes = self.generate()
+
+    def _get_class(self):
+        classes_path = os.path.expanduser(self.classes_path)
+        with open(classes_path) as f:
+            class_names = f.readlines()
+        class_names = [c.strip() for c in class_names]
+        return class_names
+
+    def _get_anchors(self):
+        anchors_path = os.path.expanduser(self.anchors_path)
+        with open(anchors_path) as f:
+            anchors = f.readline()
+            anchors = [float(x) for x in anchors.split(',')]
+            anchors = np.array(anchors).reshape(-1, 2)
+        return anchors
+
+    def generate(self):
+        model_path = os.path.expanduser(self.model_path)
+        assert model_path.endswith('.h5'), 'Keras model or weights must be a .h5 file.'
+
+        self.yolo_model = load_model(model_path, custom_objects={'Mish': Mish}, compile=False)
+
+        print('{} model, anchors, and classes loaded.'.format(model_path))
+
+        # Generate colors for drawing bounding boxes.
+        hsv_tuples = [(x / len(self.class_names), 1., 1.)
+                      for x in range(len(self.class_names))]
+        self.colors = list(map(lambda x: colorsys.hsv_to_rgb(*x), hsv_tuples))
+        self.colors = list(
+            map(lambda x: (int(x[0] * 255), int(x[1] * 255), int(x[2] * 255)),
+                self.colors))
+        np.random.seed(10101)  # Fixed seed for consistent colors across runs.
+        np.random.shuffle(self.colors)  # Shuffle colors to decorrelate adjacent classes.
+        np.random.seed(None)  # Reset seed to default.
+
+        # Generate output tensor targets for filtered bounding boxes.
+        self.input_image_shape = K.placeholder(shape=(2, ))
+        if self.gpu_num>=2:
+            self.yolo_model = multi_gpu_model(self.yolo_model, gpus=self.gpu_num)
+        boxes, scores, classes = yolo_eval(self.yolo_model.output, self.anchors,
+                len(self.class_names), self.input_image_shape,
+                score_threshold=self.score, iou_threshold=self.iou)
+        return boxes, scores, classes
+
+    def detect_image(self, image):
+
+        if self.is_fixed_size:
+            assert self.model_image_size[0]%32 == 0, 'Multiples of 32 required'
+            assert self.model_image_size[1]%32 == 0, 'Multiples of 32 required'
+            boxed_image = letterbox_image(image, tuple(reversed(self.model_image_size)))
+        else:
+            new_image_size = (image.width - (image.width % 32),
+                              image.height - (image.height % 32))
+            boxed_image = letterbox_image(image, new_image_size)
+        image_data = np.array(boxed_image, dtype='float32')
+
+        # print(image_data.shape)
+        image_data /= 255.
+        image_data = np.expand_dims(image_data, 0)  # Add batch dimension.
+
+        out_boxes, out_scores, out_classes = self.sess.run(
+            [self.boxes, self.scores, self.classes],
+            feed_dict={
+                self.yolo_model.input: image_data,
+                self.input_image_shape: [image.size[1], image.size[0]],
+                K.learning_phase(): 0
+            })
+        return_boxes = []
+        return_scores = []
+        return_class_names = []
+        for i, c in reversed(list(enumerate(out_classes))):
+            predicted_class = self.class_names[c]
+            if predicted_class != 'person':  # Modify to detect other classes.
+                continue
+            box = out_boxes[i]
+            score = out_scores[i]
+            x = int(box[1])
+            y = int(box[0])
+            w = int(box[3] - box[1])
+            h = int(box[2] - box[0])
+            if x < 0:
+                w = w + x
+                x = 0
+            if y < 0:
+                h = h + y
+                y = 0
+            return_boxes.append([x, y, w, h])
+            return_scores.append(score)
+            return_class_names.append(predicted_class)
+
+        return return_boxes, return_scores, return_class_names
+
+    def close_session(self):
+        self.sess.close()

+"""YOLO_v4 Model Defined in Keras."""
+
+from functools import wraps
+
+import numpy as np
+import tensorflow as tf
+from keras import backend as K
+from keras.engine.base_layer import Layer
+from keras.layers import Conv2D, Add, ZeroPadding2D, UpSampling2D, Concatenate, MaxPooling2D
+from keras.layers.advanced_activations import LeakyReLU
+from keras.layers.normalization import BatchNormalization
+from keras.models import Model
+from keras.regularizers import l2
+
+from yolo4.utils import compose
+
+class Mish(Layer):
+
+
+    def __init__(self, **kwargs):
+        super(Mish, self).__init__(**kwargs)
+        self.supports_masking = True
+
+    def call(self, inputs):
+        return inputs * K.tanh(K.softplus(inputs))
+
+    def get_config(self):
+        config = super(Mish, self).get_config()
+        return config
+
+    def compute_output_shape(self, input_shape):
+        return input_shape
+
+
+@wraps(Conv2D)
+def DarknetConv2D(*args, **kwargs):
+    """Wrapper to set Darknet parameters for Convolution2D."""
+    darknet_conv_kwargs = {'kernel_regularizer': l2(5e-4)}
+    darknet_conv_kwargs['padding'] = 'valid' if kwargs.get('strides')==(2,2) else 'same'
+    darknet_conv_kwargs.update(kwargs)
+    return Conv2D(*args, **darknet_conv_kwargs)
+
+def DarknetConv2D_BN_Leaky(*args, **kwargs):
+    """Darknet Convolution2D followed by BatchNormalization and LeakyReLU."""
+    no_bias_kwargs = {'use_bias': False}
+    no_bias_kwargs.update(kwargs)
+    return compose(
+        DarknetConv2D(*args, **no_bias_kwargs),
+        BatchNormalization(),
+        LeakyReLU(alpha=0.1))
+
+def DarknetConv2D_BN_Mish(*args, **kwargs):
+    """Darknet Convolution2D followed by BatchNormalization and LeakyReLU."""
+    no_bias_kwargs = {'use_bias': False}
+    no_bias_kwargs.update(kwargs)
+    return compose(
+        DarknetConv2D(*args, **no_bias_kwargs),
+        BatchNormalization(),
+        Mish())
+
+def resblock_body(x, num_filters, num_blocks, all_narrow=True):
+    '''A series of resblocks starting with a downsampling Convolution2D'''
+    # Darknet uses left and top padding instead of 'same' mode
+    preconv1 = ZeroPadding2D(((1,0),(1,0)))(x)
+    preconv1 = DarknetConv2D_BN_Mish(num_filters, (3,3), strides=(2,2))(preconv1)
+    shortconv = DarknetConv2D_BN_Mish(num_filters//2 if all_narrow else num_filters, (1,1))(preconv1)
+    mainconv = DarknetConv2D_BN_Mish(num_filters//2 if all_narrow else num_filters, (1,1))(preconv1)
+    for i in range(num_blocks):
+        y = compose(
+                DarknetConv2D_BN_Mish(num_filters//2, (1,1)),
+                DarknetConv2D_BN_Mish(num_filters//2 if all_narrow else num_filters, (3,3)))(mainconv)
+        mainconv = Add()([mainconv,y])
+    postconv = DarknetConv2D_BN_Mish(num_filters//2 if all_narrow else num_filters, (1,1))(mainconv)
+    route = Concatenate()([postconv, shortconv])
+    return DarknetConv2D_BN_Mish(num_filters, (1,1))(route)
+
+def darknet_body(x):
+    '''Darknent body having 52 Convolution2D layers'''
+    x = DarknetConv2D_BN_Mish(32, (3,3))(x)
+    x = resblock_body(x, 64, 1, False)
+    x = resblock_body(x, 128, 2)
+    x = resblock_body(x, 256, 8)
+    x = resblock_body(x, 512, 8)
+    x = resblock_body(x, 1024, 4)
+    return x
+
+def make_last_layers(x, num_filters, out_filters):
+    '''6 Conv2D_BN_Leaky layers followed by a Conv2D_linear layer'''
+    x = compose(
+            DarknetConv2D_BN_Leaky(num_filters, (1,1)),
+            DarknetConv2D_BN_Leaky(num_filters*2, (3,3)),
+            DarknetConv2D_BN_Leaky(num_filters, (1,1)),
+            DarknetConv2D_BN_Leaky(num_filters*2, (3,3)),
+            DarknetConv2D_BN_Leaky(num_filters, (1,1)))(x)
+    y = compose(
+            DarknetConv2D_BN_Leaky(num_filters*2, (3,3)),
+            DarknetConv2D(out_filters, (1,1)))(x)
+    return x, y
+
+
+def yolo4_body(inputs, num_anchors, num_classes):
+    """Create YOLO_V4 model CNN body in Keras."""
+    darknet = Model(inputs, darknet_body(inputs))
+
+    #19x19 head
+    y19 = DarknetConv2D_BN_Leaky(512, (1,1))(darknet.output)
+    y19 = DarknetConv2D_BN_Leaky(1024, (3,3))(y19)
+    y19 = DarknetConv2D_BN_Leaky(512, (1,1))(y19)
+    maxpool1 = MaxPooling2D(pool_size=(13,13), strides=(1,1), padding='same')(y19)
+    maxpool2 = MaxPooling2D(pool_size=(9,9), strides=(1,1), padding='same')(y19)
+    maxpool3 = MaxPooling2D(pool_size=(5,5), strides=(1,1), padding='same')(y19)
+    y19 = Concatenate()([maxpool1, maxpool2, maxpool3, y19])
+    y19 = DarknetConv2D_BN_Leaky(512, (1,1))(y19)
+    y19 = DarknetConv2D_BN_Leaky(1024, (3,3))(y19)
+    y19 = DarknetConv2D_BN_Leaky(512, (1,1))(y19)
+
+    y19_upsample = compose(DarknetConv2D_BN_Leaky(256, (1,1)), UpSampling2D(2))(y19)
+
+    #38x38 head
+    y38 = DarknetConv2D_BN_Leaky(256, (1,1))(darknet.layers[204].output)
+    y38 = Concatenate()([y38, y19_upsample])
+    y38 = DarknetConv2D_BN_Leaky(256, (1,1))(y38)
+    y38 = DarknetConv2D_BN_Leaky(512, (3,3))(y38)
+    y38 = DarknetConv2D_BN_Leaky(256, (1,1))(y38)
+    y38 = DarknetConv2D_BN_Leaky(512, (3,3))(y38)
+    y38 = DarknetConv2D_BN_Leaky(256, (1,1))(y38)
+
+    y38_upsample = compose(DarknetConv2D_BN_Leaky(128, (1,1)), UpSampling2D(2))(y38)
+
+    #76x76 head
+    y76 = DarknetConv2D_BN_Leaky(128, (1,1))(darknet.layers[131].output)
+    y76 = Concatenate()([y76, y38_upsample])
+    y76 = DarknetConv2D_BN_Leaky(128, (1,1))(y76)
+    y76 = DarknetConv2D_BN_Leaky(256, (3,3))(y76)
+    y76 = DarknetConv2D_BN_Leaky(128, (1,1))(y76)
+    y76 = DarknetConv2D_BN_Leaky(256, (3,3))(y76)
+    y76 = DarknetConv2D_BN_Leaky(128, (1,1))(y76)
+
+    #76x76 output
+    y76_output = DarknetConv2D_BN_Leaky(256, (3,3))(y76)
+    y76_output = DarknetConv2D(num_anchors*(num_classes+5), (1,1))(y76_output)
+
+    #38x38 output
+    y76_downsample = ZeroPadding2D(((1,0),(1,0)))(y76)
+    y76_downsample = DarknetConv2D_BN_Leaky(256, (3,3), strides=(2,2))(y76_downsample)
+    y38 = Concatenate()([y76_downsample, y38])
+    y38 = DarknetConv2D_BN_Leaky(256, (1,1))(y38)
+    y38 = DarknetConv2D_BN_Leaky(512, (3,3))(y38)
+    y38 = DarknetConv2D_BN_Leaky(256, (1,1))(y38)
+    y38 = DarknetConv2D_BN_Leaky(512, (3,3))(y38)
+    y38 = DarknetConv2D_BN_Leaky(256, (1,1))(y38)
+
+    y38_output = DarknetConv2D_BN_Leaky(512, (3,3))(y38)
+    y38_output = DarknetConv2D(num_anchors*(num_classes+5), (1,1))(y38_output)
+
+    #19x19 output
+    y38_downsample = ZeroPadding2D(((1,0),(1,0)))(y38)
+    y38_downsample = DarknetConv2D_BN_Leaky(512, (3,3), strides=(2,2))(y38_downsample)
+    y19 = Concatenate()([y38_downsample, y19])
+    y19 = DarknetConv2D_BN_Leaky(512, (1,1))(y19)
+    y19 = DarknetConv2D_BN_Leaky(1024, (3,3))(y19)
+    y19 = DarknetConv2D_BN_Leaky(512, (1,1))(y19)
+    y19 = DarknetConv2D_BN_Leaky(1024, (3,3))(y19)
+    y19 = DarknetConv2D_BN_Leaky(512, (1,1))(y19)
+
+    y19_output = DarknetConv2D_BN_Leaky(1024, (3,3))(y19)
+    y19_output = DarknetConv2D(num_anchors*(num_classes+5), (1,1))(y19_output)
+
+    yolo4_model = Model(inputs, [y19_output, y38_output, y76_output])
+
+    return yolo4_model
+
+
+def yolo_head(feats, anchors, num_classes, input_shape, calc_loss=False):
+    """Convert final layer features to bounding box parameters."""
+    num_anchors = len(anchors)
+    # Reshape to batch, height, width, num_anchors, box_params.
+    anchors_tensor = K.reshape(K.constant(anchors), [1, 1, 1, num_anchors, 2])
+
+    grid_shape = K.shape(feats)[1:3] # height, width
+    grid_y = K.tile(K.reshape(K.arange(0, stop=grid_shape[0]), [-1, 1, 1, 1]),
+        [1, grid_shape[1], 1, 1])
+    grid_x = K.tile(K.reshape(K.arange(0, stop=grid_shape[1]), [1, -1, 1, 1]),
+        [grid_shape[0], 1, 1, 1])
+    grid = K.concatenate([grid_x, grid_y])
+    grid = K.cast(grid, K.dtype(feats))
+
+    feats = K.reshape(
+        feats, [-1, grid_shape[0], grid_shape[1], num_anchors, num_classes + 5])
+
+    # Adjust preditions to each spatial grid point and anchor size.
+    box_xy = (K.sigmoid(feats[..., :2]) + grid) / K.cast(grid_shape[...,::-1], K.dtype(feats))
+    box_wh = K.exp(feats[..., 2:4]) * anchors_tensor / K.cast(input_shape[...,::-1], K.dtype(feats))
+    box_confidence = K.sigmoid(feats[..., 4:5])
+    box_class_probs = K.sigmoid(feats[..., 5:])
+
+    if calc_loss == True:
+        return grid, feats, box_xy, box_wh
+    return box_xy, box_wh, box_confidence, box_class_probs
+
+
+def yolo_correct_boxes(box_xy, box_wh, input_shape, image_shape):
+    '''Get corrected boxes'''
+    box_yx = box_xy[..., ::-1]
+    box_hw = box_wh[..., ::-1]
+    input_shape = K.cast(input_shape, K.dtype(box_yx))
+    image_shape = K.cast(image_shape, K.dtype(box_yx))
+    new_shape = K.round(image_shape * K.min(input_shape/image_shape))
+    offset = (input_shape-new_shape)/2./input_shape
+    scale = input_shape/new_shape
+    box_yx = (box_yx - offset) * scale
+    box_hw *= scale
+
+    box_mins = box_yx - (box_hw / 2.)
+    box_maxes = box_yx + (box_hw / 2.)
+    boxes =  K.concatenate([
+        box_mins[..., 0:1],  # y_min
+        box_mins[..., 1:2],  # x_min
+        box_maxes[..., 0:1],  # y_max
+        box_maxes[..., 1:2]  # x_max
+    ])
+
+    # Scale boxes back to original image shape.
+    boxes *= K.concatenate([image_shape, image_shape])
+    return boxes
+
+
+def yolo_boxes_and_scores(feats, anchors, num_classes, input_shape, image_shape):
+    '''Process Conv layer output'''
+    box_xy, box_wh, box_confidence, box_class_probs = yolo_head(feats,
+        anchors, num_classes, input_shape)
+    boxes = yolo_correct_boxes(box_xy, box_wh, input_shape, image_shape)
+    boxes = K.reshape(boxes, [-1, 4])
+    box_scores = box_confidence * box_class_probs
+    box_scores = K.reshape(box_scores, [-1, num_classes])
+    return boxes, box_scores
+
+
+def yolo_eval(yolo_outputs,
+              anchors,
+              num_classes,
+              image_shape,
+              max_boxes=200,
+              score_threshold=.5,
+              iou_threshold=.5):
+    """Evaluate YOLO model on given input and return filtered boxes."""
+    num_layers = len(yolo_outputs)
+    anchor_mask = [[6,7,8], [3,4,5], [0,1,2]]
+    input_shape = K.shape(yolo_outputs[0])[1:3] * 32
+    boxes = []
+    box_scores = []
+    for l in range(num_layers):
+        _boxes, _box_scores = yolo_boxes_and_scores(yolo_outputs[l],
+            anchors[anchor_mask[l]], num_classes, input_shape, image_shape)
+        boxes.append(_boxes)
+        box_scores.append(_box_scores)
+    boxes = K.concatenate(boxes, axis=0)
+    box_scores = K.concatenate(box_scores, axis=0)
+
+    mask = box_scores >= score_threshold
+    max_boxes_tensor = K.constant(max_boxes, dtype='int32')
+    boxes_ = []
+    scores_ = []
+    classes_ = []
+    for c in range(num_classes):
+        class_boxes = tf.boolean_mask(boxes, mask[:, c])
+        class_box_scores = tf.boolean_mask(box_scores[:, c], mask[:, c])
+        nms_index = tf.image.non_max_suppression(
+            class_boxes, class_box_scores, max_boxes_tensor, iou_threshold=iou_threshold)
+        class_boxes = K.gather(class_boxes, nms_index)
+        class_box_scores = K.gather(class_box_scores, nms_index)
+        classes = K.ones_like(class_box_scores, 'int32') * c
+        boxes_.append(class_boxes)
+        scores_.append(class_box_scores)
+        classes_.append(classes)
+    boxes_ = K.concatenate(boxes_, axis=0)
+    scores_ = K.concatenate(scores_, axis=0)
+    classes_ = K.concatenate(classes_, axis=0)
+
+    return boxes_, scores_, classes_
+
+
+def preprocess_true_boxes(true_boxes, input_shape, anchors, num_classes):
+    '''Preprocess true boxes to training input format
+
+    Parameters
+    ----------
+    true_boxes: array, shape=(m, T, 5)
+        Absolute x_min, y_min, x_max, y_max, class_id relative to input_shape.
+    input_shape: array-like, hw, multiples of 32
+    anchors: array, shape=(N, 2), wh
+    num_classes: integer
+
+    Returns
+    -------
+    y_true: list of array, shape like yolo_outputs, xywh are reletive value
+
+    '''
+    assert (true_boxes[..., 4]<num_classes).all(), 'class id must be less than num_classes'
+    num_layers = len(anchors)//3 # default setting
+    anchor_mask = [[6,7,8], [3,4,5], [0,1,2]] if num_layers==3 else [[3,4,5], [1,2,3]]
+
+    true_boxes = np.array(true_boxes, dtype='float32')
+    input_shape = np.array(input_shape, dtype='int32')
+    boxes_xy = (true_boxes[..., 0:2] + true_boxes[..., 2:4]) // 2
+    boxes_wh = true_boxes[..., 2:4] - true_boxes[..., 0:2]
+    true_boxes[..., 0:2] = boxes_xy/input_shape[::-1]
+    true_boxes[..., 2:4] = boxes_wh/input_shape[::-1]
+
+    m = true_boxes.shape[0]
+    grid_shapes = [input_shape//{0:32, 1:16, 2:8}[l] for l in range(num_layers)]
+    y_true = [np.zeros((m,grid_shapes[l][0],grid_shapes[l][1],len(anchor_mask[l]),5+num_classes),
+        dtype='float32') for l in range(num_layers)]
+
+    # Expand dim to apply broadcasting.
+    anchors = np.expand_dims(anchors, 0)
+    anchor_maxes = anchors / 2.
+    anchor_mins = -anchor_maxes
+    valid_mask = boxes_wh[..., 0]>0
+
+    for b in range(m):
+        # Discard zero rows.
+        wh = boxes_wh[b, valid_mask[b]]
+        if len(wh)==0: continue
+        # Expand dim to apply broadcasting.
+        wh = np.expand_dims(wh, -2)
+        box_maxes = wh / 2.
+        box_mins = -box_maxes
+
+        intersect_mins = np.maximum(box_mins, anchor_mins)
+        intersect_maxes = np.minimum(box_maxes, anchor_maxes)
+        intersect_wh = np.maximum(intersect_maxes - intersect_mins, 0.)
+        intersect_area = intersect_wh[..., 0] * intersect_wh[..., 1]
+        box_area = wh[..., 0] * wh[..., 1]
+        anchor_area = anchors[..., 0] * anchors[..., 1]
+        iou = intersect_area / (box_area + anchor_area - intersect_area)
+
+        # Find best anchor for each true box
+        best_anchor = np.argmax(iou, axis=-1)
+
+        for t, n in enumerate(best_anchor):
+            for l in range(num_layers):
+                if n in anchor_mask[l]:
+                    i = np.floor(true_boxes[b,t,0]*grid_shapes[l][1]).astype('int32')
+                    j = np.floor(true_boxes[b,t,1]*grid_shapes[l][0]).astype('int32')
+                    k = anchor_mask[l].index(n)
+                    c = true_boxes[b,t, 4].astype('int32')
+                    y_true[l][b, j, i, k, 0:4] = true_boxes[b,t, 0:4]
+                    y_true[l][b, j, i, k, 4] = 1
+                    y_true[l][b, j, i, k, 5+c] = 1
+
+    return y_true
+
+
+def softmax_focal_loss(y_true, y_pred, gamma=2.0, alpha=0.25):
+    """
+    Compute softmax focal loss.
+    Reference Paper:
+        "Focal Loss for Dense Object Detection"
+        https://arxiv.org/abs/1708.02002
+
+    # Arguments
+        y_true: Ground truth targets,
+            tensor of shape (?, num_boxes, num_classes).
+        y_pred: Predicted logits,
+            tensor of shape (?, num_boxes, num_classes).
+        gamma: exponent of the modulating factor (1 - p_t) ^ gamma.
+        alpha: optional alpha weighting factor to balance positives vs negatives.
+
+    # Returns
+        softmax_focal_loss: Softmax focal loss, tensor of shape (?, num_boxes).
+    """
+
+    # Scale predictions so that the class probas of each sample sum to 1
+    #y_pred /= K.sum(y_pred, axis=-1, keepdims=True)
+
+    # Clip the prediction value to prevent NaN's and Inf's
+    #epsilon = K.epsilon()
+    #y_pred = K.clip(y_pred, epsilon, 1. - epsilon)
+    y_pred = tf.nn.softmax(y_pred)
+    y_pred = tf.maximum(tf.minimum(y_pred, 1 - 1e-15), 1e-15)
+
+    # Calculate Cross Entropy
+    cross_entropy = -y_true * tf.math.log(y_pred)
+
+    # Calculate Focal Loss
+    softmax_focal_loss = alpha * tf.pow(1 - y_pred, gamma) * cross_entropy
+
+    return softmax_focal_loss
+
+
+def sigmoid_focal_loss(y_true, y_pred, gamma=2.0, alpha=0.25):
+    """
+    Compute sigmoid focal loss.
+    Reference Paper:
+        "Focal Loss for Dense Object Detection"
+        https://arxiv.org/abs/1708.02002
+
+    # Arguments
+        y_true: Ground truth targets,
+            tensor of shape (?, num_boxes, num_classes).
+        y_pred: Predicted logits,
+            tensor of shape (?, num_boxes, num_classes).
+        gamma: exponent of the modulating factor (1 - p_t) ^ gamma.
+        alpha: optional alpha weighting factor to balance positives vs negatives.
+
+    # Returns
+        sigmoid_focal_loss: Sigmoid focal loss, tensor of shape (?, num_boxes).
+    """
+    sigmoid_loss = K.binary_crossentropy(y_true, y_pred, from_logits=True)
+
+    pred_prob = tf.sigmoid(y_pred)
+    p_t = ((y_true * pred_prob) + ((1 - y_true) * (1 - pred_prob)))
+    modulating_factor = tf.pow(1.0 - p_t, gamma)
+    alpha_weight_factor = (y_true * alpha + (1 - y_true) * (1 - alpha))
+
+    sigmoid_focal_loss = modulating_factor * alpha_weight_factor * sigmoid_loss
+    #sigmoid_focal_loss = tf.reduce_sum(sigmoid_focal_loss, axis=-1)
+
+    return sigmoid_focal_loss
+
+
+def box_iou(b1, b2):
+    """
+    Return iou tensor
+
+    Parameters
+    ----------
+    b1: tensor, shape=(i1,...,iN, 4), xywh
+    b2: tensor, shape=(j, 4), xywh
+
+    Returns
+    -------
+    iou: tensor, shape=(i1,...,iN, j)
+    """
+    # Expand dim to apply broadcasting.
+    b1 = K.expand_dims(b1, -2)
+    b1_xy = b1[..., :2]
+    b1_wh = b1[..., 2:4]
+    b1_wh_half = b1_wh/2.
+    b1_mins = b1_xy - b1_wh_half
+    b1_maxes = b1_xy + b1_wh_half
+
+    # Expand dim to apply broadcasting.
+    b2 = K.expand_dims(b2, 0)
+    b2_xy = b2[..., :2]
+    b2_wh = b2[..., 2:4]
+    b2_wh_half = b2_wh/2.
+    b2_mins = b2_xy - b2_wh_half
+    b2_maxes = b2_xy + b2_wh_half
+
+    intersect_mins = K.maximum(b1_mins, b2_mins)
+    intersect_maxes = K.minimum(b1_maxes, b2_maxes)
+    intersect_wh = K.maximum(intersect_maxes - intersect_mins, 0.)
+    intersect_area = intersect_wh[..., 0] * intersect_wh[..., 1]
+    b1_area = b1_wh[..., 0] * b1_wh[..., 1]
+    b2_area = b2_wh[..., 0] * b2_wh[..., 1]
+    iou = intersect_area / (b1_area + b2_area - intersect_area)
+
+    return iou
+
+
+def box_giou(b1, b2):
+    """
+    Calculate GIoU loss on anchor boxes
+    Reference Paper:
+        "Generalized Intersection over Union: A Metric and A Loss for Bounding Box Regression"
+        https://arxiv.org/abs/1902.09630
+
+    Parameters
+    ----------
+    b1: tensor, shape=(batch, feat_w, feat_h, anchor_num, 4), xywh
+    b2: tensor, shape=(batch, feat_w, feat_h, anchor_num, 4), xywh
+
+    Returns
+    -------
+    giou: tensor, shape=(batch, feat_w, feat_h, anchor_num, 1)
+    """
+    b1_xy = b1[..., :2]
+    b1_wh = b1[..., 2:4]
+    b1_wh_half = b1_wh/2.
+    b1_mins = b1_xy - b1_wh_half
+    b1_maxes = b1_xy + b1_wh_half
+
+    b2_xy = b2[..., :2]
+    b2_wh = b2[..., 2:4]
+    b2_wh_half = b2_wh/2.
+    b2_mins = b2_xy - b2_wh_half
+    b2_maxes = b2_xy + b2_wh_half
+
+    intersect_mins = K.maximum(b1_mins, b2_mins)
+    intersect_maxes = K.minimum(b1_maxes, b2_maxes)
+    intersect_wh = K.maximum(intersect_maxes - intersect_mins, 0.)
+    intersect_area = intersect_wh[..., 0] * intersect_wh[..., 1]
+    b1_area = b1_wh[..., 0] * b1_wh[..., 1]
+    b2_area = b2_wh[..., 0] * b2_wh[..., 1]
+    union_area = b1_area + b2_area - intersect_area
+    # calculate IoU, add epsilon in denominator to avoid dividing by 0
+    iou = intersect_area / (union_area + K.epsilon())
+
+    # get enclosed area
+    enclose_mins = K.minimum(b1_mins, b2_mins)
+    enclose_maxes = K.maximum(b1_maxes, b2_maxes)
+    enclose_wh = K.maximum(enclose_maxes - enclose_mins, 0.0)
+    enclose_area = enclose_wh[..., 0] * enclose_wh[..., 1]
+    # calculate GIoU, add epsilon in denominator to avoid dividing by 0
+    giou = iou - 1.0 * (enclose_area - union_area) / (enclose_area + K.epsilon())
+    giou = K.expand_dims(giou, -1)
+
+    return giou
+
+
+def box_diou(b1, b2):
+    """
+    Calculate DIoU loss on anchor boxes
+    Reference Paper:
+        "Distance-IoU Loss: Faster and Better Learning for Bounding Box Regression"
+        https://arxiv.org/abs/1911.08287
+
+    Parameters
+    ----------
+    b1: tensor, shape=(batch, feat_w, feat_h, anchor_num, 4), xywh
+    b2: tensor, shape=(batch, feat_w, feat_h, anchor_num, 4), xywh
+
+    Returns
+    -------
+    diou: tensor, shape=(batch, feat_w, feat_h, anchor_num, 1)
+    """
+    b1_xy = b1[..., :2]
+    b1_wh = b1[..., 2:4]
+    b1_wh_half = b1_wh/2.
+    b1_mins = b1_xy - b1_wh_half
+    b1_maxes = b1_xy + b1_wh_half
+
+    b2_xy = b2[..., :2]
+    b2_wh = b2[..., 2:4]
+    b2_wh_half = b2_wh/2.
+    b2_mins = b2_xy - b2_wh_half
+    b2_maxes = b2_xy + b2_wh_half
+
+    intersect_mins = K.maximum(b1_mins, b2_mins)
+    intersect_maxes = K.minimum(b1_maxes, b2_maxes)
+    intersect_wh = K.maximum(intersect_maxes - intersect_mins, 0.)
+    intersect_area = intersect_wh[..., 0] * intersect_wh[..., 1]
+    b1_area = b1_wh[..., 0] * b1_wh[..., 1]
+    b2_area = b2_wh[..., 0] * b2_wh[..., 1]
+    union_area = b1_area + b2_area - intersect_area
+    # calculate IoU, add epsilon in denominator to avoid dividing by 0
+    iou = intersect_area / (union_area + K.epsilon())
+
+    # box center distance
+    center_distance = K.sum(K.square(b1_xy - b2_xy), axis=-1)
+    # get enclosed area
+    enclose_mins = K.minimum(b1_mins, b2_mins)
+    enclose_maxes = K.maximum(b1_maxes, b2_maxes)
+    enclose_wh = K.maximum(enclose_maxes - enclose_mins, 0.0)
+    # get enclosed diagonal distance
+    enclose_diagonal = K.sum(K.square(enclose_wh), axis=-1)
+    # calculate DIoU, add epsilon in denominator to avoid dividing by 0
+    diou = iou - 1.0 * (center_distance) / (enclose_diagonal + K.epsilon())
+
+    # calculate param v and alpha to extend to CIoU
+    #v = 4*K.square(tf.math.atan2(b1_wh[..., 0], b1_wh[..., 1]) - tf.math.atan2(b2_wh[..., 0], b2_wh[..., 1])) / (math.pi * math.pi)
+    #alpha = v / (1.0 - iou + v)
+    #diou = diou - alpha*v
+
+    diou = K.expand_dims(diou, -1)
+    return diou
+
+
+def _smooth_labels(y_true, label_smoothing):
+    label_smoothing = K.constant(label_smoothing, dtype=K.floatx())
+    return y_true * (1.0 - label_smoothing) + 0.5 * label_smoothing
+
+
+def yolo4_loss(args, anchors, num_classes, ignore_thresh=.5, label_smoothing=0, use_focal_loss=False, use_focal_obj_loss=False, use_softmax_loss=False, use_giou_loss=False, use_diou_loss=False):
+    '''Return yolo4_loss tensor
+
+    Parameters
+    ----------
+    yolo_outputs: list of tensor, the output of yolo_body or tiny_yolo_body
+    y_true: list of array, the output of preprocess_true_boxes
+    anchors: array, shape=(N, 2), wh
+    num_classes: integer
+    ignore_thresh: float, the iou threshold whether to ignore object confidence loss
+
+    Returns
+    -------
+    loss: tensor, shape=(1,)
+
+    '''
+    num_layers = len(anchors)//3 # default setting
+    yolo_outputs = args[:num_layers]
+    y_true = args[num_layers:]
+    anchor_mask = [[6,7,8], [3,4,5], [0,1,2]] if num_layers==3 else [[3,4,5], [0,1,2]]
+    input_shape = K.cast(K.shape(yolo_outputs[0])[1:3] * 32, K.dtype(y_true[0]))
+    grid_shapes = [K.cast(K.shape(yolo_outputs[l])[1:3], K.dtype(y_true[0])) for l in range(num_layers)]
+    loss = 0
+    total_location_loss = 0
+    total_confidence_loss = 0
+    total_class_loss = 0
+    m = K.shape(yolo_outputs[0])[0] # batch size, tensor
+    mf = K.cast(m, K.dtype(yolo_outputs[0]))
+
+    for l in range(num_layers):
+        object_mask = y_true[l][..., 4:5]
+        true_class_probs = y_true[l][..., 5:]
+        if label_smoothing:
+            true_class_probs = _smooth_labels(true_class_probs, label_smoothing)
+
+        grid, raw_pred, pred_xy, pred_wh = yolo_head(yolo_outputs[l],
+             anchors[anchor_mask[l]], num_classes, input_shape, calc_loss=True)
+        pred_box = K.concatenate([pred_xy, pred_wh])
+
+        # Darknet raw box to calculate loss.
+        raw_true_xy = y_true[l][..., :2]*grid_shapes[l][::-1] - grid
+        raw_true_wh = K.log(y_true[l][..., 2:4] / anchors[anchor_mask[l]] * input_shape[::-1])
+        raw_true_wh = K.switch(object_mask, raw_true_wh, K.zeros_like(raw_true_wh)) # avoid log(0)=-inf
+        box_loss_scale = 2 - y_true[l][...,2:3]*y_true[l][...,3:4]
+
+        # Find ignore mask, iterate over each of batch.
+        ignore_mask = tf.TensorArray(K.dtype(y_true[0]), size=1, dynamic_size=True)
+        object_mask_bool = K.cast(object_mask, 'bool')
+        def loop_body(b, ignore_mask):
+            true_box = tf.boolean_mask(y_true[l][b,...,0:4], object_mask_bool[b,...,0])
+            iou = box_iou(pred_box[b], true_box)
+            best_iou = K.max(iou, axis=-1)
+            ignore_mask = ignore_mask.write(b, K.cast(best_iou<ignore_thresh, K.dtype(true_box)))
+            return b+1, ignore_mask
+        _, ignore_mask = tf.while_loop(lambda b,*args: b<m, loop_body, [0, ignore_mask])
+        ignore_mask = ignore_mask.stack()
+        ignore_mask = K.expand_dims(ignore_mask, -1)
+
+        if use_focal_obj_loss:
+            # Focal loss for objectness confidence
+            confidence_loss = sigmoid_focal_loss(object_mask, raw_pred[...,4:5])
+        else:
+            confidence_loss = object_mask * K.binary_crossentropy(object_mask, raw_pred[...,4:5], from_logits=True)+ \
+                (1-object_mask) * K.binary_crossentropy(object_mask, raw_pred[...,4:5], from_logits=True) * ignore_mask
+
+        if use_focal_loss:
+            # Focal loss for classification score
+            if use_softmax_loss:
+                class_loss = softmax_focal_loss(true_class_probs, raw_pred[...,5:])
+            else:
+                class_loss = sigmoid_focal_loss(true_class_probs, raw_pred[...,5:])
+        else:
+            if use_softmax_loss:
+                # use softmax style classification output
+                class_loss = object_mask * K.expand_dims(K.categorical_crossentropy(true_class_probs, raw_pred[...,5:], from_logits=True), axis=-1)
+            else:
+                # use sigmoid style classification output
+                class_loss = object_mask * K.binary_crossentropy(true_class_probs, raw_pred[...,5:], from_logits=True)
+
+
+        if use_giou_loss:
+            # Calculate GIoU loss as location loss
+            raw_true_box = y_true[l][...,0:4]
+            giou = box_giou(pred_box, raw_true_box)
+            giou_loss = object_mask * box_loss_scale * (1 - giou)
+            giou_loss = K.sum(giou_loss) / mf
+            location_loss = giou_loss
+        elif use_diou_loss:
+            # Calculate DIoU loss as location loss
+            raw_true_box = y_true[l][...,0:4]
+            diou = box_diou(pred_box, raw_true_box)
+            diou_loss = object_mask * box_loss_scale * (1 - diou)
+            diou_loss = K.sum(diou_loss) / mf
+            location_loss = diou_loss
+        else:
+            # Standard YOLO location loss
+            # K.binary_crossentropy is helpful to avoid exp overflow.
+            xy_loss = object_mask * box_loss_scale * K.binary_crossentropy(raw_true_xy, raw_pred[...,0:2], from_logits=True)
+            wh_loss = object_mask * box_loss_scale * 0.5 * K.square(raw_true_wh-raw_pred[...,2:4])
+            xy_loss = K.sum(xy_loss) / mf
+            wh_loss = K.sum(wh_loss) / mf
+            location_loss = xy_loss + wh_loss
+
+        confidence_loss = K.sum(confidence_loss) / mf
+        class_loss = K.sum(class_loss) / mf
+        loss += location_loss + confidence_loss + class_loss
+        total_location_loss += location_loss
+        total_confidence_loss += confidence_loss
+        total_class_loss += class_loss
+
+    # Fit for tf 2.0.0 loss shape
+    loss = K.expand_dims(loss, axis=-1)
+
+    return loss #, total_location_loss, total_confidence_loss, total_class_loss
+
+
+def yolo_loss(args, anchors, num_classes, ignore_thresh=.5, print_loss=False):
+    '''Return yolo_loss tensor
+
+    Parameters
+    ----------
+    yolo_outputs: list of tensor, the output of yolo_body or tiny_yolo_body
+    y_true: list of array, the output of preprocess_true_boxes
+    anchors: array, shape=(N, 2), wh
+    num_classes: integer
+    ignore_thresh: float, the iou threshold whether to ignore object confidence loss
+
+    Returns
+    -------
+    loss: tensor, shape=(1,)
+
+    '''
+    num_layers = len(anchors)//3 # default setting
+    yolo_outputs = args[:num_layers]
+    y_true = args[num_layers:]
+    anchor_mask = [[6,7,8], [3,4,5], [0,1,2]] if num_layers==3 else [[3,4,5], [1,2,3]]
+    input_shape = K.cast(K.shape(yolo_outputs[0])[1:3] * 32, K.dtype(y_true[0]))
+    grid_shapes = [K.cast(K.shape(yolo_outputs[l])[1:3], K.dtype(y_true[0])) for l in range(num_layers)]
+    loss = 0
+    m = K.shape(yolo_outputs[0])[0] # batch size, tensor
+    mf = K.cast(m, K.dtype(yolo_outputs[0]))
+
+    for l in range(num_layers):
+        object_mask = y_true[l][..., 4:5]
+        true_class_probs = y_true[l][..., 5:]
+
+        grid, raw_pred, pred_xy, pred_wh = yolo_head(yolo_outputs[l],
+             anchors[anchor_mask[l]], num_classes, input_shape, calc_loss=True)
+        pred_box = K.concatenate([pred_xy, pred_wh])
+
+        # Darknet raw box to calculate loss.
+        raw_true_xy = y_true[l][..., :2]*grid_shapes[l][::-1] - grid
+        raw_true_wh = K.log(y_true[l][..., 2:4] / anchors[anchor_mask[l]] * input_shape[::-1])
+        raw_true_wh = K.switch(object_mask, raw_true_wh, K.zeros_like(raw_true_wh)) # avoid log(0)=-inf
+        box_loss_scale = 2 - y_true[l][...,2:3]*y_true[l][...,3:4]
+
+        # Find ignore mask, iterate over each of batch.
+        ignore_mask = tf.TensorArray(K.dtype(y_true[0]), size=1, dynamic_size=True)
+        object_mask_bool = K.cast(object_mask, 'bool')
+        def loop_body(b, ignore_mask):
+            true_box = tf.boolean_mask(y_true[l][b,...,0:4], object_mask_bool[b,...,0])
+            iou = box_iou(pred_box[b], true_box)
+            best_iou = K.max(iou, axis=-1)
+            ignore_mask = ignore_mask.write(b, K.cast(best_iou<ignore_thresh, K.dtype(true_box)))
+            return b+1, ignore_mask
+        _, ignore_mask = K.control_flow_ops.while_loop(lambda b,*args: b<m, loop_body, [0, ignore_mask])
+        ignore_mask = ignore_mask.stack()
+        ignore_mask = K.expand_dims(ignore_mask, -1)
+
+        # K.binary_crossentropy is helpful to avoid exp overflow.
+        xy_loss = object_mask * box_loss_scale * K.binary_crossentropy(raw_true_xy, raw_pred[...,0:2], from_logits=True)
+        wh_loss = object_mask * box_loss_scale * 0.5 * K.square(raw_true_wh-raw_pred[...,2:4])
+        confidence_loss = object_mask * K.binary_crossentropy(object_mask, raw_pred[...,4:5], from_logits=True)+ \
+            (1-object_mask) * K.binary_crossentropy(object_mask, raw_pred[...,4:5], from_logits=True) * ignore_mask
+        class_loss = object_mask * K.binary_crossentropy(true_class_probs, raw_pred[...,5:], from_logits=True)
+
+        xy_loss = K.sum(xy_loss) / mf
+        wh_loss = K.sum(wh_loss) / mf
+        confidence_loss = K.sum(confidence_loss) / mf
+        class_loss = K.sum(class_loss) / mf
+        loss += xy_loss + wh_loss + confidence_loss + class_loss
+        if print_loss:
+            loss = tf.Print(loss, [loss, xy_loss, wh_loss, confidence_loss, class_loss, K.sum(ignore_mask)], message='loss: ')
+    return loss

+"""Miscellaneous utility functions."""
+
+from functools import reduce
+
+from PIL import Image
+import numpy as np
+from matplotlib.colors import rgb_to_hsv, hsv_to_rgb
+
+def compose(*funcs):
+    """Compose arbitrarily many functions, evaluated left to right.
+
+    Reference: https://mathieularose.com/function-composition-in-python/
+    """
+    # return lambda x: reduce(lambda v, f: f(v), funcs, x)
+    if funcs:
+        return reduce(lambda f, g: lambda *a, **kw: g(f(*a, **kw)), funcs)
+    else:
+        raise ValueError('Composition of empty sequence not supported.')
+
+def letterbox_image(image, size):
+    '''resize image with unchanged aspect ratio using padding'''
+    iw, ih = image.size
+    w, h = size
+    scale = min(w/iw, h/ih)
+    nw = int(iw*scale)
+    nh = int(ih*scale)
+
+    image = image.resize((nw,nh), Image.BICUBIC)
+    new_image = Image.new('RGB', size, (128,128,128))
+    new_image.paste(image, ((w-nw)//2, (h-nh)//2))
+    return new_image
+
+def rand(a=0, b=1):
+    return np.random.rand()*(b-a) + a
+
+def get_random_data(annotation_line, input_shape, random=True, max_boxes=100, jitter=.3, hue=.1, sat=1.5, val=1.5, proc_img=True):
+    '''random preprocessing for real-time data augmentation'''
+    line = annotation_line.split()
+    image = Image.open(line[0])
+    iw, ih = image.size
+    h, w = input_shape
+    box = np.array([np.array(list(map(int,box.split(',')))) for box in line[1:]])
+
+    if not random:
+        # resize image
+        scale = min(w/iw, h/ih)
+        nw = int(iw*scale)
+        nh = int(ih*scale)
+        dx = (w-nw)//2
+        dy = (h-nh)//2
+        image_data=0
+        if proc_img:
+            image = image.resize((nw,nh), Image.BICUBIC)
+            new_image = Image.new('RGB', (w,h), (128,128,128))
+            new_image.paste(image, (dx, dy))
+            image_data = np.array(new_image)/255.
+
+        # correct boxes
+        box_data = np.zeros((max_boxes,5))
+        if len(box)>0:
+            np.random.shuffle(box)
+            if len(box)>max_boxes: box = box[:max_boxes]
+            box[:, [0,2]] = box[:, [0,2]]*scale + dx
+            box[:, [1,3]] = box[:, [1,3]]*scale + dy
+            box_data[:len(box)] = box
+
+        return image_data, box_data
+
+    # resize image
+    new_ar = w/h * rand(1-jitter,1+jitter)/rand(1-jitter,1+jitter)
+    scale = rand(.25, 2)
+    if new_ar < 1:
+        nh = int(scale*h)
+        nw = int(nh*new_ar)
+    else:
+        nw = int(scale*w)
+        nh = int(nw/new_ar)
+    image = image.resize((nw,nh), Image.BICUBIC)
+
+    # place image
+    dx = int(rand(0, w-nw))
+    dy = int(rand(0, h-nh))
+    new_image = Image.new('RGB', (w,h), (128,128,128))
+    new_image.paste(image, (dx, dy))
+    image = new_image
+
+    # flip image or not
+    flip = rand()<.5
+    if flip: image = image.transpose(Image.FLIP_LEFT_RIGHT)
+
+    # distort image
+    hue = rand(-hue, hue)
+    sat = rand(1, sat) if rand()<.5 else 1/rand(1, sat)
+    val = rand(1, val) if rand()<.5 else 1/rand(1, val)
+    x = rgb_to_hsv(np.array(image)/255.)
+    x[..., 0] += hue
+    x[..., 0][x[..., 0]>1] -= 1
+    x[..., 0][x[..., 0]<0] += 1
+    x[..., 1] *= sat
+    x[..., 2] *= val
+    x[x>1] = 1
+    x[x<0] = 0
+    image_data = hsv_to_rgb(x) # numpy array, 0 to 1
+
+    # correct boxes
+    box_data = np.zeros((max_boxes,5))
+    if len(box)>0:
+        np.random.shuffle(box)
+        box[:, [0,2]] = box[:, [0,2]]*nw/iw + dx
+        box[:, [1,3]] = box[:, [1,3]]*nh/ih + dy
+        if flip: box[:, [0,2]] = w - box[:, [2,0]]
+        box[:, 0:2][box[:, 0:2]<0] = 0
+        box[:, 2][box[:, 2]>w] = w
+        box[:, 3][box[:, 3]>h] = h
+        box_w = box[:, 2] - box[:, 0]
+        box_h = box[:, 3] - box[:, 1]
+        box = box[np.logical_and(box_w>1, box_h>1)] # discard invalid box
+        if len(box)>max_boxes: box = box[:max_boxes]
+        box_data[:len(box)] = box
+
+    return image_data, box_data

+absl-py==0.9.0
+asgiref==3.2.7
+astor==0.8.1
+astroid==2.3.3
+cachetools==4.1.0
+certifi==2020.4.5.1
+chardet==3.0.4
+colorama==0.4.3
+cycler==0.10.0
+Django==3.0.5
+et-xmlfile==1.0.1
+gast==0.2.2
+google-auth==1.14.1
+google-auth-oauthlib==0.4.1
+google-pasta==0.2.0
+grpcio==1.28.1
+h5py==2.10.0
+idna==2.9
+image==1.5.31
+isort==4.3.21
+jdcal==1.4.1
+joblib==0.14.1
+Keras==2.3.1
+Keras-Applications==1.0.8
+Keras-Preprocessing==1.1.0
+kiwisolver==1.2.0
+lazy-object-proxy==1.4.3
+Markdown==3.2.1
+matplotlib==3.2.1
+mccabe==0.6.1
+numpy==1.18.3
+oauthlib==3.1.0
+opencv-python==4.2.0.34
+openpyxl==3.0.3
+opt-einsum==3.2.1
+pandas==1.0.3
+Pillow==7.1.1
+protobuf==3.11.3
+pyasn1==0.4.8
+pyasn1-modules==0.2.8
+pylint==2.4.4
+pyparsing==2.4.7
+python-dateutil==2.8.1
+pytz==2019.3
+PyYAML==5.3.1
+requests==2.23.0
+requests-oauthlib==1.3.0
+rsa==4.0
+scikit-learn==0.22.2.post1
+scipy==1.4.1
+seaborn==0.10.1
+six==1.14.0
+sklearn==0.0
+sqlparse==0.3.1
+tensorboard==1.15.0
+tensorflow==1.15.2
+tensorflow-estimator==1.15.1
+tensorflow-gpu==1.15.2
+tensorflow-gpu-estimator==2.1.0
+termcolor==1.1.0
+typed-ast==1.4.1
+urllib3==1.25.9
+Werkzeug==1.0.1
+wrapt==1.11.2
+xlrd==1.2.0

@@ -6,12 +6,14 @@
     "start": "node ./bin/www"
   },
   "dependencies": {
+    "child_process": "^1.0.2",
     "cookie-parser": "~1.4.4",
     "debug": "~2.6.9",
     "ejs": "~2.6.1",
     "express": "~4.16.1",
     "http-errors": "~1.6.3",
     "morgan": "~1.9.1",
-    "multer": "^1.4.2"
+    "multer": "^1.4.2",
+    "python-shell": "^2.0.1"
   }
 }

@@ -2,10 +2,13 @@ var express = require('express');
 var router = express.Router();
 var multer = require("multer");
 var path = require("path");
+var PythonShell = require('python-shell');
+var spawn = require("child_process").spawn;
+
 
 var storage = multer.diskStorage({
     destination: function(req, file, callback) {
-        callback(null, "upload/")
+        callback(null, "../deep_sort_yolov4/")
     },
     filename: function(req, file, callback) {
         var extension = path.extname(file.originalname);
@@ -21,20 +24,33 @@ var upload = multer({
 
 // 뷰 페이지 경로
 router.get('/', function(req, res, next) {
-    res.render("index")
+    res.render('index', { title: 'Express' });
 });
 
 // 2. 파일 업로드 처리
-router.post('/create', upload.single("File"), async(req, res) => {
+router.post('/create', upload.single("File"), function(req, res) {
     // 3. 파일 객체
-    var file = req.file
-
-    // 4. 파일 정보
-    var result = {
-        originalName: file.originalname,
-        size: file.size,
+    var file = req.file;
+    var options = {
+        mode: 'text',
+        pythonPath: __dirname + '/../../venv/Scripts/python.exe',
+        pythonOptions: ['-u'],
+        scriptPath: __dirname + '/../../deep_sort_yolov4/',
+        args: [file.originalname]
+    };
+    var shell1 = new PythonShell.PythonShell('main.py', options);
+    shell1.end(function(err) {
+        if (err) throw err;
+        else {
+            res.render('result', { file_name: file.originalname.split('.')[0] });
         }
-
-    res.json(result);
+    });
+    // PythonShell.PythonShell.run('main.py', options, (err, results) => {
+    //     if (err) throw err;
+    //     PythonShell.PythonShell.run('kmean.py', options2, (err, results) => {
+    //         if (err) throw err;
+    //         res.render('result', { file_name: file.originalname.split('.')[0] });
+    //     });
+    // });
 });
 module.exports = router;
\ No newline at end of file

@@ -32,7 +32,8 @@
                 <div class="sidebar-brand-icon rotate-n-15">
                     <i class="fas fa-thumbs-up"></i>
                 </div>
-                <div class="sidebar-brand-text mx-3">유동 인구 분석</div>
+                <div class="sidebar-brand-text mx-3">유동 인구 분석
+                </div>
             </a>
 
             <!-- Divider -->

@@ -86,10 +86,10 @@
                 <!-- /.container-fluid -->
                 <div class="container row">
                     <div class="container-video col-md-6">
-                        <video src="data/output2.avi" width="400" controls autoplay></video>
+                        <video src="data/<%=file_name%>.mp4" width="400" controls autoplay></video>
                     </div>
                     <div class="container-kmeans col-md-6">
-                        <img src="data/test3_kmeans.png" style="width: 100%;" alt="Kmeans Image">
+                        <img src="data/<%=file_name%>_kmeans.png" style="width: 100%;" alt="Kmeans Image">
                     </div>
                 </div>
                 <div class="container-kibana"></div>
@@ -127,9 +127,6 @@
     <!-- Custom scripts for all pages-->
     <script src="javascripts/sb-admin-2.min.js"></script>
 
-    <!-- Page level plugins -->
-    <script src="vendor/chart.js/Chart.min.js"></script>
-
     <!-- Page level custom scripts -->
     <!-- <script src="javascripts/demo/chart-pie-demo.js"></script> -->
 </body>