Researches of scale adaptive Mean Shift target tracking algorithm

Fuzzy Threshold Segmentation has been used for target tracking to improve deficiency that the bandwidth of kernel function of Mean Shift is not changeable. The weight of gradation transformation is ascertained by the contrast of similarity of RGB component between the target and background. According to certain rules,the bandwidth of kernel function is corrected with the statistic of target pixels. The scale adaptive of track window has been realized. Experiment results show the new algorithm can track objects accurately, effectively and real time. The new algorithm also overcomes the limitations of the traditional one.