Feature point matching of curved surface and robust uncertainty

1 .INTRODUCTIONStereovisionis well used in wide aspects of robotvision, object recognition1 ~7], VR8]and recon-struction9]and so on. Stereo matching is a hardapproachin stereovision especially for curved sur-face due to occlusions and noises . There were al-ready some related methods1 ~3]such as matchingunder robust esti mation of epipolar geometryrecovery1]or selecting salient points fromi magepair2]. Muchi mportance has been attached to theuncertainty of the corner position and i…

Feature point matching of curved surface and robust uncertainty

Stereo matching is an important research area in stereovision and stereo matching of curved surface is especially crucial. A nov el correspondence algorithm is presented and its matching uncertainty is compute d robustly for feature points of curved surface. The corners are matched by usin g homography constraint besides epipolar constraint to solve the occlusion probl em. The uncertainty sources are analyzed. A cost function is established and act s as an optimal rule to compute the matching uncertainty. An adaptive scheme Gau ss weights are put forward to make the matching results robust to noises. It mak es the practical application of corner matching possible. From the experimental results of an image pair of curved surface it is shown that computing uncertaint y robustly can restrain the affection caused by noises to the matching precision .