摘要
立体匹配是计算机视觉研究中的关键问题。相比于双目视觉,三目视觉能够获得更多的信息和额外的极线约束消除立体匹配的歧义性。为了提高三目立体匹配的精度,提出一种基于自适应权值和视差校准的三目立体匹配方法。将双目视觉中有效的自适应权值窗选择算法应用到三目视觉中,进行匹配窗的选择;提出一种新的目标图像选择算法,能够合理利用平行基线三目立体视觉系统中不同目标图像提供的信息,有效地消除遮挡,提高匹配的精度;提出一种适用于三目视觉的视差校准算法,利用三目图像像素间的色彩相似性和距离约束将初始匹配的视差结果进行校准,得到最终的视差图。实验结果表明,本文算法结构简单,能够生成浓密、高精度的视差图。
Stereo-matching is the most important problem in computer vision. Comparing with the binocular vision, trinocular vision can provide more information and an additional constraint of epipolar line to eliminate the ambiguities of stereo-matching. A trinocular stereo-matching algorithm based on adaptive support-weight and disparity adjustment has been proposed to improve the accuracy of trinocular vision. Adaptive support-weight which is an efficient window selection method in binocular stereo is applied in the trinocular stereo. A robust target image selection method is also proposed for using the information of different target images in trinocular stereo-vision system efficiently. So occlusion can be handled well by this means. Furthermore, a trinocular disparity adjustment method is introduced. By using the constraint of color similarity and space proximity, this method can correct many errors of initial disparity. The experimental results show that the proposed algorithm can extract accurate disparity. It＇s concise and efficient.
出处
《光学学报》
EI
CAS
CSCD
北大核心
2008年第4期734-738,共5页
Acta Optica Sinica
基金
国家自然科学基金(60527001)资助课题
关键词
机器视觉
三目立体匹配
自适应权值
目标图像选择
视差校准
machine vision
trinocular stereo-matching
adaptive support-weight
target image selection
disparity adjustment
