基于各向异性核函数的均值漂移跟踪算法

均值漂移算法是一种将迭代轨迹滑向局部邻域内均值的迭代算法，已应用于目标跟踪领域。传统的均值漂移算法通常采用各向同性核函数进行跟踪，但视频序列中的跟踪目标的结构随时间而变化，尤其当目标结构快速变化时，基于各向同性核函数的均值漂移跟踪算法常常会导致目标的丢失。该文采用各向异性核函数均值漂移算法实现目标跟踪，由于该核函数的形状、大小、方向能自适应于目标局部结构的变化，保证了跟踪效果的稳定性和鲁棒性。实验结果证明该算法是有效的。

Object Tracking by Anisotropic Kernel Mean Shift

Mean shift, an iterative procedure that shifts each data point to the average of data points in its neighborhood, has been applied to object tracking. However, with the changing structure of object in video sequences, traditional mean shift tracker by isotropic kernel often loses the object, especially when object structure varies fast. This paper implements object tracking with anisotropic kernel mean shift in which the shape, scale, and orientation of the kernels adapt to the changing object structure. The algorithm ensures tracking robust and real-time. Experimental results show it is effective.