Improved belief propagation with istinctiveness measure for stereo matching

Abstract

Stereo matching is one of the most active research areas in computer vision. Many algorithms including local algorithms and global algorithms have been proposed. As a global algorithm, belief propagation (BP) is one of the best algorithms for stereo matching. But BP algorithm is still a difficult problem because of ambiguous matching of the edge points. In this work, the priori on both matching image pairs appearances is considered during the matching process. And the distinctiveness of points, which is the difference between points in the reference image and target image, is introduced to reflect the prior of edge points appearances. Experimental results show that the proposed method performs better than the conventional BP algorithm.