Title :
Affine parameter estimation from the trace transform
Author :
Kadyrov, Alexander ; Petrou, Maria
Author_Institution :
Sch. of Electron., Comput. & Math., Surrey Univ., Guildford, UK
Abstract :
We assume that we are given two images, one of which is an affine version of the other, and present a way of computing the parameters of the affine transform between the two images. We present a generic theoretical framework in terms of which other proposed methods can be interpreted, and show how in this framework one can recover the affine parameters in a way that is robust to various effects, like occlusion and illumination variation.
Keywords :
Radon transforms; feature extraction; image reconstruction; parameter estimation; affine parameter estimation; affine version; circus functions; generic theoretical framework; illumination variation; invariant feature construction; occlusion; trace transform; two images; Computed tomography; Fourier transforms; Image reconstruction; Iterative methods; Mathematics; Parameter estimation; Robustness;
Conference_Titel :
Pattern Recognition, 2002. Proceedings. 16th International Conference on
Print_ISBN :
0-7695-1695-X
DOI :
10.1109/ICPR.2002.1048423