A Stereo Matching Algorithm with Support Regions Based on Color and Texture Estimate
Cheng Hui Zhu, Qi Jiao, Jian‐Ping Wang, Xiao Bing Xu · Advanced materials research · 2012
A stereo matching algorithm with support regions based on color and texture estimate is proposed. Firstly, the initial support regions are selected from the image according to the distribution of the quantized color labels. Then, the texture similarity is used to determine the arm length growing and combine adjacent regions. The accurate support regions are obtained. Thirdly, the support weight is introduced under the constraint of support region. Finally, the initial disparity can be corrected by using disparity adjustment method until a reasonable disparity map is obtained. The experimental results show that the good disparity result can be obtained.