Semiglobal Stereo Matching Algorithm Based on Improved Census Transformation

Binocular stereo matching is in an effort to obtain disparity maps from binocular images taken by the same camera. Existing methods resulted in mismatches in the discontinuous regions. In this study, we put forward a innovative binocular stereo correspondence method based on an improved census transformation. After that, semiglobal stereo matching algorithm was performed on matching cost aggregation stage. Finally, we further conducted Left and Right Consistency Check and median filter to generate better disparities. The results indicated that compared with the original census transform, the disparities obtained by our method was continuous and dense, and the mismatch rate is lower.