This paper describes a novel technique to recover large similarity
transformations (rotation/scale/translation) and moderate perspective
deformations among image pairs. We introduce a hybrid algorithm that
features log-polar mappings and nonlinear least squares optimization.
The use of log-polar techniques in the spatial domain is introduced
as a preprocessing module to recover large scale changes (e.g., at
least four-fold) and arbitrary rotations. Although log-polar techniques
are used in the Fourier-Mellin transform to accommodate rotation
and scale in the frequency domain, its use in registering images
subjected to very large scale changes has not yet been exploited
in the spatial domain. In this paper, we demonstrate the superior
performance of the log-polar transform in featureless image registration
in the spatial domain. We achieve subpixel accuracy through the use
of nonlinear least squares optimization. The registration process
yields the eight parameters of the perspective transformation that
best aligns the two input images. Extensive testing was performed
on uncalibrated real images and an array of 10,000 image pairs with
known transformations derived from the Corel Stock Photo Library
of royalty-free photographic images.
%0 Journal Article
%1 ZokaiOct.2005
%A Zokai, S.
%A Wolberg, G.
%D Oct. 2005
%K Corel Fourier Fourier-Mellin Levenberg-Marquardt Library, Photo Stock approximations, array domain image image, large-scale least least-squares log-polar mappings, nonlinear optimisation, optimization, pair, photographic photography projective registration, similarity, spatial squares transform, transformation, transforms,
%N 10
%P 1422-1434
%R 10.1109/TIP.2005.854501
%T Image registration using log-polar mappings for recovery of large-scale
similarity and projective transformations
%V 14
%X This paper describes a novel technique to recover large similarity
transformations (rotation/scale/translation) and moderate perspective
deformations among image pairs. We introduce a hybrid algorithm that
features log-polar mappings and nonlinear least squares optimization.
The use of log-polar techniques in the spatial domain is introduced
as a preprocessing module to recover large scale changes (e.g., at
least four-fold) and arbitrary rotations. Although log-polar techniques
are used in the Fourier-Mellin transform to accommodate rotation
and scale in the frequency domain, its use in registering images
subjected to very large scale changes has not yet been exploited
in the spatial domain. In this paper, we demonstrate the superior
performance of the log-polar transform in featureless image registration
in the spatial domain. We achieve subpixel accuracy through the use
of nonlinear least squares optimization. The registration process
yields the eight parameters of the perspective transformation that
best aligns the two input images. Extensive testing was performed
on uncalibrated real images and an array of 10,000 image pairs with
known transformations derived from the Corel Stock Photo Library
of royalty-free photographic images.
@article{ZokaiOct.2005,
abstract = { This paper describes a novel technique to recover large similarity
transformations (rotation/scale/translation) and moderate perspective
deformations among image pairs. We introduce a hybrid algorithm that
features log-polar mappings and nonlinear least squares optimization.
The use of log-polar techniques in the spatial domain is introduced
as a preprocessing module to recover large scale changes (e.g., at
least four-fold) and arbitrary rotations. Although log-polar techniques
are used in the Fourier-Mellin transform to accommodate rotation
and scale in the frequency domain, its use in registering images
subjected to very large scale changes has not yet been exploited
in the spatial domain. In this paper, we demonstrate the superior
performance of the log-polar transform in featureless image registration
in the spatial domain. We achieve subpixel accuracy through the use
of nonlinear least squares optimization. The registration process
yields the eight parameters of the perspective transformation that
best aligns the two input images. Extensive testing was performed
on uncalibrated real images and an array of 10,000 image pairs with
known transformations derived from the Corel Stock Photo Library
of royalty-free photographic images.},
added-at = {2011-03-27T19:35:34.000+0200},
author = {Zokai, S. and Wolberg, G.},
biburl = {https://www.bibsonomy.org/bibtex/2f893cfcb3899068a93d26f411a317db3/cocus},
doi = {10.1109/TIP.2005.854501},
file = {:./01510678.pdf:PDF},
interhash = {2335faa31c347d0fd457a1b062cbf556},
intrahash = {f893cfcb3899068a93d26f411a317db3},
issn = {1057-7149},
journaltitle = {#ieeetip#},
keywords = {Corel Fourier Fourier-Mellin Levenberg-Marquardt Library, Photo Stock approximations, array domain image image, large-scale least least-squares log-polar mappings, nonlinear optimisation, optimization, pair, photographic photography projective registration, similarity, spatial squares transform, transformation, transforms,},
number = 10,
pages = { 1422-1434},
timestamp = {2011-03-27T19:35:45.000+0200},
title = {Image registration using log-polar mappings for recovery of large-scale
similarity and projective transformations},
volume = 14,
year = {Oct. 2005}
}