Stereo matching based on phase congruency corner detector

Feature point matching is the most common one among all kinds of stereo matching. However, since feature points are unique, the disparity map through feature matching is also sparse. In this paper, proposed a dense disparity estimation method that combines the reliability of feature-based correspondence methods and a reliable feature operator. The new operator uses the principal moments of the phase congruency information to determine corner information. The resulting corner operator is highly localized and has responses that are invariant to image contrast. This results in reliable feature detection under varying illumination conditions with fixed thresholds. Selecting those feature points that allow left-right correspondence based on phase correlation surrounding each point. And use the sparse correspondences at feature points as a constraint to control the computation of dense disparity via regularized block matching that minimizes matching and disparity smoothness errors. Experimental results show that this method can eliminate many kinds of outliers effectively.