Premium
Matching of uncalibrated stereo images by elastic deformation
Author(s) -
Li WanChiu,
Leung C. H.,
Hung Y. S.
Publication year - 2004
Publication title -
international journal of imaging systems and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.359
H-Index - 47
eISSN - 1098-1098
pISSN - 0899-9457
DOI - 10.1002/ima.20024
Subject(s) - epipolar geometry , outlier , artificial intelligence , computer science , computer vision , matching (statistics) , deformation (meteorology) , computation , point set registration , function (biology) , point (geometry) , line (geometry) , fundamental matrix (linear differential equation) , image (mathematics) , mathematics , algorithm , geometry , mathematical analysis , geography , statistics , biology , meteorology , evolutionary biology
We propose a method for uncalibrated stereo matching. The method applies gradual elastic deformation to the line segments in a pair of images until they match with each other. By using an energy function and a neighborhood function, matching is performed in a coarse‐to‐fine manner. Our method gives point correspondences with a low proportion of outliers and is robust in the uncalibrated case (with no need to estimate the epipolar geometry). The computation complexity is proportional to the square of the number of line segments in the images, which is relatively efficient compared with other elaborate methods. © 2005 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 14, 198–205, 2004; Published online in Wiley Inter‐Science (www.interscience.wiley.com). DOI 10.1002/ima.20024