一种快速的视频跟踪算法

A Fast Video Tracking Algorithm

在摄像头固定的视频监控系统中，本文给出了一种基于最小化随机复杂度的快速有效的视频目标跟踪算法。该算法结合了基于背景减技术和目标相似性建模的优势，不需要调整参数权重。实验结果表明，应用本方法能快速的估计运动目标的位置和矩形轮廓。

In the video surveillance system with fixed camera,this paper puts forward a quick and effective video target tracking algorithm based on the minimization of stochastic complexity.This algorithm combines the advantages of the minus technology based on the background and modeling of target similarity without need for adjusting the parameter weight.The experimental results show that this method can estimate the location and the rectangular shape of the moving target quickly.