A stereo video object segmentation algorithm based on motion detection and disparity

With the development of the stereo video, the technology of the stereo video object segmentation has gradually become a research hotpot. In this paper, a stereo video object segmentation algorithm based on motion detection and disparity is proposed, which utilizes the information of motion, disparity and edging features to extract moving objects from video sequences. Firstly, the eight-neighbor motion detection is used to get the change detection mask (CDM) of moving objects. Secondly, the stereo matching is adopted to get the disparity map which will then be used to modify the CDM. Finally, the edging information and mathematic morphology post-processing are applied to get the accurate video object. Experimental results demonstrate that the proposed algorithm can successfully extract moving objects from stereo video sequences whose background have other moving objects.