Matching methods for oblique stereo images on the basis of geometry constraint

作者：Yao Guo Biao^*; Deng Ka Zhong; Yang Hua Chao; Du Quan Ye

来源：Journal of China University of Mining & Technology, 2013, 42(4): 676-682.

• Affine transformations
• Exterior orientation
• Geometry constraints
• Least-squares matching
• Normalized cross correlation
• Projective distortions
• Stereo-image
• Turning points

This study proposed a precise matching algorithm using geometry constraint for oblique images on the basis of severe projective distortions and scene complexity. The proposed algorithm includes following two stages: 1) space coordinates were calculated by correspondents of nadir-view stereo images and exterior orientation elements, and then correspondents on squint image were matched by making full use of matches on one of nadir-view images;according to elevation of correspondent turned successfully, affine approximate planes were divided, and meanwhile coarse matches were classified;2) precise matching based on local affine constrain. According to each cluster of coarse matching pairs, firstly, seed points were selected and sequenced, and the seed pair was matched by least squares matching method, then affine transformation matrix A could be obtained, further coarse matching pairs located in neighborhood of seed point were matched precisely by adaptive normalized cross correlation on the basis of A. Experimental results show that the proposed algorithm would provide satisfactory achievement.

• 出版日期2013
• 单位中国矿业大学（北京）