科研之友
微信
新浪微博
Facebook
分享链接
摘要
A new color space feature based particle filtering tracking algorithm is proposed. The algorithm uses K-means clustering for object description and EMD for the distance measure between two clustering. For the problem of the clustering algorithm's time-consuming, GPU is used to parallel accelerating. The experimental results confirm that the proposed algorithm can yield good results as compared with a color histogram-based- method and the processing speed upgrades great are increased dramatically.
