摘要
提出了一种新的快速立体匹配算法。该算法根据图像边缘特征自适应地变化窗口,并采用灰度差平方和(SSD)作为相似判定准则计算初始视差图;通过左右一致性校验去除误匹配点,在构造能量函数时,将初始视差作为能量函数的一个参考项;最后采用图切割算法求取使全局能量最小的视差最优分配。用标准图像测试了所提出的方法,并与其他方法进行了比较。实验结果表明,该算法不仅能够保留基于图切割的立体匹配算法能较好地处理大的低纹理区域和遮挡像素的优点,而且匹配时间短,运行时间比原有算法缩短了约2/3。
According to the disadvantage of large computation of stereo matching algorithm based on graph cuts, a new stereo matching algorithm is presented. By changing adaptively the window based on image edge features, the SSD matching method is used as a similarity decision criterion to determine the initial disparity map. Then, a left-right conformity verification is adopted to remove mistake matching points, when the energy function is constructed, the initial disparity is used as a reference of function to get the optimization of disparity distribution of the minimum global energy by graph cuts. The proposed algorithm is tested with standard stereo images and compared with other methods, experimental results show that proposed algorithm not only has better performance for processing the large low texture areas and the occluded pixels, but also has shorter matching time, which is two-thirds shorter than that of original algorithm.
出处
《光学精密工程》
EI
CAS
CSCD
北大核心
2008年第6期1117-1121,共5页
Optics and Precision Engineering
基金
国家自然科学基金资助项目(No.6037062)
关键词
立体匹配
自适应窗口
视差图
图切割
能量函数
stereo matching
adaptive window
disparity map
graph cuts
energy function
