A dynamic programming approach for stereo line matching with structural information

In this paper, dynamic programming approach is used to solve the correspondence problem in stereo vision. A new structural weight which defines the similarity of the structure between the left and right image, is added so as to improve the accuracy of the matching process. The proposed method uses local structural information so as to reduce the errors of mismatch due to imperfect feature extraction such as missing line and broken line. The algorithm has been tested on real scenes to confirm the usefulness of the proposed method