Active target tracking based on Mean Shift algorithm and color template matching
Zhenhong Shang, Jinxia Shang, Xiaofang Cai, Liu Hui, Guofeng Shu, Siyang Yu, Zhenping Qiang · 2011
An active tracking algorithm synthesized Mean Shift Algorithm and Color template matching was proposed in dynamic background. After changing of camera perspective or focus, the algorithm, based on the fact: color template matching with the moving target is independent of the position in frames, locates target in new frame using color template matching and then tracks target using Mean Shift algorithm, by initializing Mean Shift search windows with the new location of target. The algorithm can track moving target correctly and keep tracking target in the center of the frame. Experiments show that the algorithm is great of robustness and practical.