基于颜色纹理直方图的带权分块均值漂移目标跟踪算法

Weighted Fragments-Based Meanshift Tracking Using Color-Texture Histogram

融合了传统的颜色直方图并基于局部二元模式表示的纹理特征来表示跟踪目标,提出一种基于颜色纹理直方图的带权分块均值漂移目标跟踪算法.为了更好地解决目标遮挡和姿势改变的问题,在跟踪过程中将跟踪目标分割成多个互不遮挡的矩形分块,对每一个矩形分块独立采用基于背景权重的均值漂移算法,并结合每一个分块求得的最佳目标位置得到目标物体在下一帧中的位置.实验结果表明,该算法和传统的基于颜色直方图特征的均值漂移算法相比具有更准确的跟踪结果,并且对于目标遮挡和姿势改变的视频序列具有更强的鲁棒性.

A weighted fragment-based meanshift tracking algorithm using color-texture histogram is proposed in this paper.In addition to using traditional color histogram features,the texture features of the object are also extracted by using the local binary pattern（LBP） to represent the target object.In order to handle the problems of partial occlusions and pose changes,we propose a weighted fragment-based meanshift tracking algorithm.For an efficient tracking,we represent the target with multiple image fragments and use independent background-weighted meanshift tracker for each fragment.Finally we combine all the tracking results of the fragments to get the target position in the next frame.Experimental results show that the proposed algorithm obtains more accurate tracking results and is better accounted for partial occlusions and pose changes when compared with the basic meanshift tracker based on traditional color histogram features.