以常见的监控摄像头的实时视频信号为输入采用基于神经网络的YOLO目标检测算法进行检测与特征提取，最后利用自己的多目标追踪算法实现人流量统计和目标对象保存(Taking the real-time video signal of the surveillance camera as the input, the YOLO target detection algorithm based on the neural network is used to detect and extract the feature. Finally, the multi-target tracking algorithm is used to realize the statistics of the traffic statistics and the target object)