Robust Fast Tracking Algorithm Based on Mean Shift
Tracking objects in video uLsing the mean shift technique has been the SUbject of considerable attention.A problem with MS is that jt becomes easy for the tracker to get confused with background which having the similar histogram distribution.We proposed a robust fast tracking algorithm Which is based on the adaptive over-relaxed strategy.Meanwhile the relation between the bandwidth and the mean-shift vectors was derived.A novel tracker designed by this theory is robust to the background Which having the similar histogram distribution.Compared with the standard MS algorithm on real application data sets.the proposed method decreases the time of iterations required to achieve convergence and increase the robustness.A physical simulation system for tracking objects Was designed and realized by employing DM642EVM as video signal processor.TMS320LF2407 as the two axis servo system controller
MeanShill Physical Simulation Tracking moving object
Jian Sun Zheng You Feng-qi Zhou Jun Zhou
国际会议
The International Conference Information Computing and Automation(2007国际信息计算与自动化会议)
成都
英文
385-387
2007-12-19（万方平台首次上网日期，不代表论文的发表时间）