一种结合ORB特征匹配的均值漂移目标跟踪算法

针对传统的mean-shift跟踪算法基于单一颜色特征空间,在复杂背景下难以对目标进行准确跟踪这问题,提出了一种结合ORB特征匹配的mean-shift目标跟踪算法。该算法在mean-shift算法的基础上利用改进的ORB特征匹配算法修正目标跟踪窗口并实时更新目标特征模板,通过计算前后两帧图像中目标中心的欧式距离与色彩模板的巴氏距离来判定跟踪是否失败,当目标跟踪失败时,不改变目标模板,继续搜索下一帧图像中的目标。实验结果表明,与均值漂移算法和基于其他同类特征的改进算法相比,该算法提高了在复杂背景下目标跟踪的精度,并能满足实时性要求。

A kind of mean-shift target tracking algorithm combine with ORB feature matching

In view of the traditional mean-shift tracking algorithm based on single color feature space, and it is difficult to accurately track the target in complex background, and a mean-shift target tracking algorithm based on ORB feature matching is putted forward. The algorithm on the basis of mean shift algorithm and using the improved orb feature matching algorithm to modified object tracking window and update the target feature template timely. Determination tracking failure through calculate Euclidean distance of the center of the target and color template Bhattacharyya distance of the target in adjacent two frames. When the target tracking failure, it does not change the target template and continue to search for target in the next frame image. The experimental results show that, compared with mean shift algorithm and other improved algorithms based on similar features, the proposed algorithm can improve the accuracy of target tracking in complex background, and can meet the requirement of real time