Home // ICCGI 2013, The Eighth International Multi-Conference on Computing in the Global Information Technology // View article

Global Illumination-Invariant Fast Sub-Pixel Image Registration

Authors:
Andrew Gilman
Arno Leist

Keywords: image registration; sub-pixel; direct; least-squares optimal; linear computational complexity

Abstract:
Sub-pixel image registration is an important part of many image processing and computer vision applications. We propose a computationally simple direct (i.e., non-iterative) method for sub-pixel registration of translated images. To register two images, first a global space-invariant resampling filter is computed that is a least-squares optimal predictor of one image from the pixel values of the other image. Then, the coefficients of this filter are linearly combined to compute the offset between the two images. The computational cost of this algorithm is linear in the number of pixels. The accuracy and efficiency of the proposed algorithm is demonstrated to be better than a range of existing methods for images with various levels of high-frequency detail and at various noise levels.

Pages: 95 to 100

Copyright: Copyright (c) IARIA, 2013

Publication date: July 21, 2013

Published in: conference

ISSN: 2308-4529

ISBN: 978-1-61208-283-7

Location: Nice, France

Dates: from July 21, 2013 to July 26, 2013