期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2012
卷号：1
期号：4
页码：649-652
出版社：Shri Pannalal Research Institute of Technolgy
摘要：The multi moving targets tracking are a typical job in the field of visual surveillance. The main difficulties in targets tracking are fast motion of the target, suddenly velocity variations, clutters, complex object shapes and occlusions, etc. Target tracking are basically explain in to three main parts: point tracking, kernel tracking and silhouette tracking. Here we can track multiple targets with the help of mean shift algorithm. For tracking purpose, the mean shift method requires a kernel that mean's a circle or a rectangle or an ellipse region. First with the help of mean shift, we can evaluate the objects locations. In mean shift algorithm the Bhattacharyya coefficient show an important task for tracking. It is a similarity function which can be obtained by target model and target candidate. The experimental results describe that the mean shift algorithm is practical for tracking point of view.
