基于颜色和LBP多特征的mean shift的跟踪算法

Mean shift target tracking based on color and LBP features

mean shift目标跟踪使用任一个单独特征都很难对大幅度的光照、背景变化和车辆大范围运动具有强鲁棒性,如单一的颜色特征对光照较为敏感,光照突变容易导致目标丢失。当背景颜色和目标颜色相近时也易造成目标丢失,因此利用目标的单一特征信息往往很难实现鲁棒的目标跟踪。文章提出基于颜色和LBP多特征mean shift跟踪方法,使跟踪结果不再过分依赖某一特征,增强了对背景变化、目标大范围运动的鲁棒性。

In mean shift target tracking,if a separate feature is used,it is very difficult to achieve strong robustness to substantial illumination,background changes or vehicle movement in large range.For example,the color feature is sensitive to light,so the light mutation easily leads to the target losing. When the background color and the target color are similar,the target is also easily lost.So it is often difficult to achieve robust target tracking by using single target feature information.In this paper,the mean shift tracking approach based on color and local binary pattern（LBP）features is proposed, which makes the tracking results no longer rely too much on one feature,and can enhance the robustness to background changes and large-range motion.