先用背景差分完成目标提取，将运动目标的相关信息放入到链表中，通过帧差完成对多目标的跟踪，达到实时性的要求
运动目标的相关信息存放于结构体中，可以在此基础上判断目标的动作行为跟踪基本框架，opencv编写，利用背景差分，质心法可以跟踪多个目标，并分配ID号，记录行动轨迹，可用于车辆道路等，大家可以参考学习下，效果有待改进，不同的视频会有不同的效果。(The relevant information of the moving object is stored in the structure. Based on this, we can judge the basic fr a mework of the moving behavior tracking of the target. After opencv writing, background difference and centroid method can be used to track multiple targets and allocate ID numbers to record the moving trajectory. In vehicles such as roads, we can learn by reference, the effect needs to be improved, different videos will have different effects.)