Symmetric Pixel-Group Based Stereo Matching for Occlusion Handling

In this paper, we propose a symmetric pixel-group stereo model for handling occlusion in a segment-based style. Firstly, both images are segmented based on color, disparity, and the segments of the other image sequencely. Then the uniqueness constraint is embodied in pixel-group level. Finally, a symmetric belief propagation (BP) optimization framework is used to find correspondence and occlusions simultaneously. Results obtained for benchmark indicate that the proposed method is able to compete with the state-of-the-art algorithms.