Disparity Selective Stereo Matching Using Correlation Confidence Measure

Test code to reproduce the results of the paper.<br>This work presents a robust stereo matching method for occluded regions. First, we generate cost volumes using the CEN-SUS transform and the scale-invariant feature transform(SIFT). Then, label-based cost volumes are aggregated using adaptive support weight and SLIC scheme from generated two cost volumes. In order to obtain optimal disparity by two label-based cost volumes, we select the disparity corresponding to high confidence similarity of CENSUS or SIFT with minimum cost point.

用于复现该论文实验结果的测试代码。本研究提出了一种针对遮挡区域的鲁棒立体匹配方法。首先，我们采用CEN-SUS变换与尺度不变特征变换（Scale-Invariant Feature Transform，SIFT）生成代价体。随后，基于生成的两类代价体，我们采用自适应支持权重与简单线性迭代聚类（Simple Linear Iterative Clustering，SLIC）方案对基于标签的代价体进行聚合。为了基于两类基于标签的代价体获取最优视差，我们选取与CENSUS或SIFT具备高置信度相似度且代价最小的点所对应的视差。