Saliency Cut in Stereo Images

In this paper, we propose a novel saliency-aware stereo images segmentation approach using the high-order energy items, which utilizes the disparity map and statistical information of stereo images to enrich the high-order potentials. To the best of our knowledge, our approach is first one to formulate the automatic stereo cut as the high-order energy optimization problems, which simultaneously segments the foreground objects in left and right images using the proposed high-order energy function. The relationships of stereo correspondence by disparity maps are further employed to enhance the connections between the left and right images. Experimental results demonstrate that the proposed approach can effectively improve the saliency-aware segmentation performance of stereo images.