In order to solve the problem of low matching accuracy in stereo matching algorithm,based on the analysis of the optical feature in RGB color image,triple normal of pixels in a 3-dimensional RGB image space is proposed, which can reflect the high frequency information of the RGB image planes,respectively. In order to get the accurate dense disparity map between the stereo pairs,based on the adaptive support weight approach in RGB vector space,a matching algorithm,which combines the RGB normal similarities to compute the corresponding support weights and dis-similarity measurements,is proposed. Testing by the Middlebury stereo benchmark,the result of the proposed algo-rithm shows more accurate disparity than many state-of-the-art stereo matching algorithms.%为了解决立体匹配算法中匹配精度不高的问题,在分析RGB彩色空间图像的光学特性的基础上,提出了RGB彩色空间R、G、B三彩色分量亮度法线的概念,并得出亮度法线反映了RGB图像像素三分量的高频信息的结论.为了获得双目立体图像间的精确稠密视差图,提出了采用RGB三分量亮度法线相似度来计算自适应权值的局部立体匹配算法,通过Middlebury测试平台与其他当前流行的立体匹配算法进行结果比较,实验结果显示,提出的算法得到了更加精确的匹配结果.
展开▼
