Stereo matching based on adaptive support-weight approach in RGB vector space.
Stereo matching based on adaptive support-weight approach in RGB vector space.
复制标题
DOI：
10.1364/ao.51.003538
复制
发表时间：
2012-06
期刊：
影响因子：
1.9
通讯作者：
Yingnan Geng;Yan Zhao;Hexin Chen
中科院分区：
文献类型：
--
作者：
Yingnan Geng;Yan Zhao;Hexin Chen
梯度相似性是一种简单但功能强大的数据描述符，在立体匹配中表现出健壮性。本文定义了一种用于立体匹配的RGB向量空间。在自适应支持权方法的基础上，提出了一种利用RGB向量空间中的像素梯度相似度、颜色相似度和贴近度来计算相应的支持权和相异度量的匹配算法。在Middlebury立体声基准上的实验结果表明，该算法的性能优于其他的立体匹配算法，并且具有梯度相似性的算法在立体匹配中取得了更好的效果。
Gradient similarity is a simple, yet powerful, data descriptor which shows robustness in stereo matching. In this paper, a RGB vector space is defined for stereo matching. Based on the adaptive support-weight approach, a matching algorithm, which uses the pixel gradient similarity, color similarity, and proximity in RGB vector space to compute the corresponding support-weights and dissimilarity measurements, is proposed. The experimental results are evaluated on the Middlebury stereo benchmark, showing that our algorithm outperforms other stereo matching algorithms and the algorithm with gradient similarity can achieve better results in stereo matching.