We propose a new dense local stereo matching framework for gray-level images based on an adaptive local segmentation using a\r\ndynamic threshold.We define a new validity domain of the frontoparallel assumption based on the local intensity variations in the\r\n4 neighborhoods of the matching pixel. The preprocessing step smoothes low-textured areas and sharpens texture edges, whereas\r\nthe postprocessing step detects and recovers occluded and unreliable disparities. The algorithm achieves high stereo reconstruction\r\nquality in regions with uniform intensities as well as in textured regions. The algorithm is robust against local radiometrical\r\ndifferences and successfully recovers disparities around the objects edges, disparities of thin objects, and the disparities of the\r\noccluded region.Moreover, our algorithm intrinsically prevents errors caused by occlusion to propagate into nonoccluded regions.\r\nIt has only a small number of parameters. The performance of our algorithm is evaluated on theMiddlebury test bed stereo images.\r\nIt ranks highly on the evaluation list outperforming many local and global stereo algorithms using color images. Among the local\r\nalgorithms relying on the frontoparallel assumption, our algorithm is the best-ranked algorithm. We also demonstrate that our\r\nalgorithm is working well on practical examples as for disparity estimation of a tomato seedling and a 3D reconstruction of a face.
Loading....