基于Mean Shift行人跟踪算法的改进

Improvement of Pedestrian Tracking Algorithm Based on Mean Shift

构建了一种Mean Shift跟踪算法,通过梯度方向直方图的描述,搭建了一种Mean Shift核心算法,算法的核心是通过迭代以达到快速收敛.通过构建的算法,基本实现了梯度方向直方图为根本的Mean Shift跟踪算法对灰暗图像中的行人进行有效的跟踪.

In this paper, a Mean Shift tracking algorithm is constructed. Through the descrip- tion of gradient direction histogram, a core algorithm of Mean Shift is built, the core of the al- gorithm is to achieve fast convergence by iteration. By constructing the algorithm, we imple- ment the Mean Shift tracking algorithm based on the gradient direction histogram to track the pedestrian in the gray image effectively.