DIRECT LINEAR SUB-PIXEL CORRELATION BY INCORPORATION OF NEIGHBOR PIXELS INFORMATION AND ROBUST ESTIMATION OF WINDOW TRANSFORMATION

Authors
Citation
Zd. Lan et R. Mohr, DIRECT LINEAR SUB-PIXEL CORRELATION BY INCORPORATION OF NEIGHBOR PIXELS INFORMATION AND ROBUST ESTIMATION OF WINDOW TRANSFORMATION, Machine vision and applications, 10(5-6), 1998, pp. 256-268
Citations number
28
Categorie Soggetti
Computer Science Cybernetics","Computer Science Artificial Intelligence","Engineering, Eletrical & Electronic","Computer Science Artificial Intelligence","Computer Science Cybernetics
ISSN journal
09328092
Volume
10
Issue
5-6
Year of publication
1998
Pages
256 - 268
Database
ISI
SICI code
0932-8092(1998)10:5-6<256:DLSCBI>2.0.ZU;2-9
Abstract
Standard methods for sub-pixel matching are iterative and nonlinear; t hey are also sensitive to false initialization and window deformation. In this paper, we present a linear method that incorporates informati on from neighboring pixels. Two algorithms are presented: one 'fast' a nd one 'robust'. They both start from an initial rough estimate of the matching. The fast one is suitable for pairs of images requiring negl igible window deformation. The robust method is slower but more genera l and more precise. It eliminates false matches in the initialization by using robust estimation of the local affine deformation. The first algorithm attains an accuracy of 0.05 pixels for interest points and 0 .06 for random points in the translational case. For the general case, if the deformation is small, the second method gives an accuracy of 0 .05 pixels; while for large deformation, it gives an accuracy of about 0.06 pixels for points of interest and 0.10 pixels for random points. They are very few false matches in all cases, even if there are many in the initialization.