计算机技术与发展2016，Vol.26Issue(10)：55-59,5.DOI:10.3969/j.issn.1673-629X.2016.10.012
基于局部特征匹配的双焦单目立体视觉深度估计
Depth Estimation for Bifocal Monocular Stereo Vision Based on Local Image Feature Descriptors Matching
摘要
Abstract
Focused on the issue of the focal length change of image based on the bifocal monocular stereo vision and the high rate of false matching of similar image,the use of local feature descriptors and‘Two-step Matching Method’ for image matching is proposed. Local feature descriptors are used to complete the image feature matching in the monocular stereo vision system and compared with each other.‘Two-step matching method’ is used to obtain the set of feature points,and two sets of feature points are obtained by changing the order of two images,the small focal length one and the large focal length one,and then a new set can be got by an intersection operation be-tween the above two point sets,thereby the depth estimation computation could be completed by the new set. Experimental results show that the SSURF has the fastest matching rate,and the depth value obtained by the above method is approximately equivalent to the ideal depth value,so it verifies that the local feature descriptors used for image matching based on bifocal monocular stereo vision to compute the depth estimation is feasible.关键词
双焦成像/图像匹配/局部特征/深度估计/SSURFKey words
bifocal imaging/image matching/local feature/depth estimation/SSURF分类
信息技术与安全科学引用本文复制引用
冯春,吴小锋,尹飞鸿,杨名利..基于局部特征匹配的双焦单目立体视觉深度估计[J].计算机技术与发展,2016,26(10):55-59,5.基金项目
江苏省自然科学基金青年项目(BK20140252) （BK20140252）
江苏省高校自然科学研究面上资助项目(14KJB510003) （14KJB510003）
江苏省教育厅自然科学基金资助项目(13KJD460003) （13KJD460003）
