A key issue in performing stereo reconstruction is to find stereo correspondences. For this purpose a novel unified framework for arbitrary area based correlation methods is presented. The commonly neglected problem of ambiguities is addressed and a solution based on relaxation of match candidates is introduced. The method comes with the major benefits of a reverse matching step at almost no extra computational cost. The results of one match direction is used to select match candidates. Then these candidates are correlated backward with a different method. The proposed framework does not require any initial estimations. Experiments are performed on ground truth under varying geometric and varying noise conditions. Finally, the method is compared to a number of well known techniques.

Perspective distortion, occlusion and specular reflection are challenging problems in shape-from-stereo. In this paper we review one recently published area-based stereo matching algorithm (Bhat and Nayar, 1998) designed to be robust in these cases. Although the algorithm is an important contribution to stereo-matching, we show that its coefficient has a low discriminatory power, which leads to a significant number of multiple best matches. In order to cope with this drawback we introduce a new normalized ordinal correlation coefficient. Experiments showing the behavior of the proposed coefficient are performed on various datasets including real data with ground truth. The new coefficient reduces the occurrence of multiple best matches to almost zero per cent. It also shows a more robust and equally accurate behavior. These benefits are achieved at almost no additional computational costs.