期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2015
卷号：4
期号：7
页码：3099-3104
出版社：Shri Pannalal Research Institute of Technolgy
摘要：The aim of object tracking is to find moving object in a frames extracted from video input. A object tracking system consist of three steps of data tracking; object extraction, object recognition and tracking. This framework takes the region of target object as input in the previous frame and a stored feature of the target object, and locates the object in the current frame. It finds the frame region that matches the input. As the object changes its view, the object model is updated. For tracking of the objects Local Steering Kernels can be used and for matching variant mask template matching can be used. The proposed method overcomes the problem of occlusion.
