基于视频的车辆检测与跟踪算法研究

Vehicle Detection and Tracking Based on Video

基于视频的车辆检测以及车辆跟踪是智能交通系统中的重要部分。本文在混合高斯背景模型的基础上,通过差分法分割出检测目标,利用检测目标的位置信息和色彩信息,找到与之最匹配的目标轨迹,从而实现车辆的跟踪。实验表明,该方法具有很高的检测与跟踪效率,同时能够满足智能交通系统的适时性要求。

Vehicle detection and tracking based on video is an important part of intellect traffic system. On the purpose of finding the best matching trajectory, this paper proposed a method that utilizes both position information and color information of detected object through background subtraction, based on Gaussian mixture model. Experimental result shows that method possesses higher efficiency in vehicle detection and tracking.