一种灰度成像扩展目标跟踪方法

灰度图像色彩信息贫乏会导致易陷于局部相似，使跟踪点发生漂移。针对此问题，提出了基于目标成像空间域与特征域分布的改进直方图模式扩展目标跟踪算法。以基于像素位置和像素灰度值的直方图模式作为目标模式；相似度测量采用基于Bhattacharyya系数定义的距离；用均值偏移迭代进行匹配区域搜索；通过建立卡尔曼预测跟踪策略解决目标被遮挡时的跟踪问题。试验结果表明，采用该算法，跟踪点漂移由原来十几个像素的波动减少到仅偶有1个像素的抖动。

Tracking algorithm for extended target of gray imaging

An improved algorithm based on histogram for tracking extended target is proposed in order to overcome the problem of tracking point shift, which comes from partial similarity between target and its background due to lack of color information in the gray images. The new algorithm is set up based on the histogram of the coordinates and pixel gray scale. Similarity measurement is determined by the distance defined by Bhattacharyya coefficient and target is located through mean shift. A prediction based on a Kalman filter is adopted to solve the problems when the target is under occlusion. Test results show that the modified algorithm can significantly improve the tracking stability for gray targets. With the modified algorithm, the tracking point fluctuation could be reduced to only 1 pixel, sporadically during test.