Stereo Matching Algorithm Based on Improved Census Transform
Zhifang Yang, Zilong Li · 2023
Aiming at the problem that the traditional Census transform is too dependent on the center pixel, which is easily affected by noise, this paper proposed a stereo matching algorithm based on improved Census transform. The weighted average calculation is introduced into the traditional Census transform. The weights of the pixels in the window are reasonably assigned, and the weighted average of all the pixels in the window is used to replace the central pixel to calculate the matching cost, which improves the robustness of the algorithm to noise. The Winner-Take-All algorithm is used to calculate the matching cost. Finally, the left-right consistency check and median filtering are used to complete the disparity optimization. The proposed algorithm is tested on the data set provided by the public test platform. The results show that the proposed algorithm improves the matching accuracy compared with the original algorithm.