针对动态规划匹配算法的误匹配与狭窄遮挡物问题,提出一种在轮廓图中提取背景控制点的立体匹配算法,在立体图像对的轮廓图中选择背景控制点,利用动态规划在视差空间图像中搜索最优路径,根据视差约束以及狭窄遮挡物的判定公式完善视差图。仿真实验结果表明,该算法能够降低在视差不连续区域匹配上的误匹配率。
Aiming at the problem of error matching and straitness occluding in dynamic plan matching algorithm, a stereo matching algorithm for contour map ground control points extraction is proposed. It chooses the ground control points in virtue of contour perceiving. The dynamic plan is used to search the best path in disparity space image. The disparity image is perfected based on the constraint of disparity image and the determinant formula of straitness occluding. Simulation experimental results show this algorithm can reduce the false matching rate in disparity's discontinuity areas matching.