摘要
交通目标的检测是智能交通系统(ITS)中的一项关键技术,基于视频跟踪方法的检测技术是目前研究的热点。介绍了近年来提出的一些主要的基于视频的运动目标跟踪方法,对各种方法进行了归类,并分析比较了这些方法的优缺点。在此基础上,着重介绍了一种快速运动目标跟踪方法——MeanShift算法。该算法主要利用图像的颜色统计直方图作为特征,利用Bhattacharyya距离作为目标匹配相似性测度,采用梯度优化方法完成对运动目标的快速跟踪。该方法非常适合对交通目标的跟踪。
The detection of traffic target is a key technology in the intelligent transportation systems(ITS),and at present,the detection technology based on video tracking is a research topic of much interest.Some of the major video-based tracking methods about moving targets in recent years are introduced in this paper,which are classified and analyzed the advantages and disadvantages.On this basis,focus on a fast moving target tracking method-MeanShift algorithm.The algorithm mainly uses the color curve and surface of the image as the main feature,and use Bhattacharyya distance as a similarity measure standard to match the target,and at last use gradient optimization method to complete the fast tracking of moving targets.This method is suitable for tracking targets on traffic.
出处
《计算机技术与发展》
2010年第7期44-47,共4页
Computer Technology and Development
基金
西安市科技攻关项目(GG06014)