自适应分块颜色直方图的MeanShift跟踪算法

传统颜色直方图的MeanShift(MS)算法只考虑了目标颜色的统计信息,不包含目标的空间信息,当目标颜色与背景颜色相近时,容易导致不准确跟踪或跟踪丢失。针对该问题,提出了一种自适应空间颜色直方图的MeanShift跟踪算法。该算法根据目标对象的最新外接矩形尺寸,确定对象分块方法,根据各块的Bhattacharyya系数值,确定各块的权重系数。其中,自适应分块的颜色直方图包含了自适应分块方法和目标的空间信息;加权Bhattacharyya系数考虑到不同块对整体相似度的不同影响。实验表明,文中算法采用自适应分块方法和加权Bhattacharyya系数法,比传统的MS算法和固定分块的MS算法具有更好的跟踪性能。

MeanShift Tracking Algorithm with Adaptive Block Color Histogram

Traditional color histogram MeanShift(MS) algorithm only considered object’s color statistical information,and didn’t contain object’s space information,so when the object color closed to the background color,the traditional MS algorithm easily caused object’s tracking inaccurately or lost.Aimmed at this issue,a MeanShift tracking algorithm with adaptive block color histogram was proposed in this article,which determined block method by the size of the lastest enclosing rectangle and determined their weight coefficient by the Bhattacharyya coefficient of each block.Among them,the adaptive block color histogram contained adaptive block method and object’s spatial information,and the weighted Bhattacharyya coefficient considered different block’s influence to the overall similarity.Experimental results show that the proposed method which uses the adaptive block method and weighted Bhattachcharyya coefficient method is better than the traditional MS algorithm and fixed block MS algorithm in tracking performance.