A motion field reconstruction scheme for smooth boundary video object segmentation

Motion segmentation is a classic and on-going research topic which is an important pre-stage for many video processes. The reliability of the motion field calculation directly determines either success or failure of such segmentation. Due to the aperture property of optical flow calculation, motion estimation at moving object boundary remains a challenging ill-posed problem. In this paper, we propose a reliable opticalflow estimation method from the view of missing data reconstruction. Furthermore, to overcome the occlusion problem at motion boundaries, we apply a motion segmentation scheme which integrates with spatial segmentation. Experimental results of proposed method and other's are presented. I. INTRODUCTON Motion segmentation is the process of categorizing pixels based on the homogeneities of their motions. It has been used for, but not limited to, layered representation of video clips, video object definition to improve compression efficiency, video surveillance, etc. As ...

• Type: Conference
• Year: 2004
• Where: ICIP
• Authors: Jean Gao, Ninad Thakoor, Sungyong Jung
• Keywords: Affine Motion Model, Approach Yields Motion, Dominant Motion Approach, ICIP 2004, Image Processing, Motion Segmentation, Motion Segmentation Scheme