Super-resolution from unregistered omnidirectional images

This paper addresses the problem of super- resolution from low resolution spherical images that are not perfectly registered. Such a problem is typically en- countered in omnidirectional vision scenarios with re- duced resolution sensors in imperfect settings. Several spherical images with arbitrary rotations in the SO(3) rotation group are used for the reconstruction of higher resolution images. We first describe the impact of the registration error on the Spherical Fourier Transform coefficients. Then, we formulate the joint registration and reconstruction problem as a least squares norm minimization problem in the transform domain. Exper- imental results show that the proposed scheme leads to effective approximations of the high resolution images, even with large registration errors. The quality of the reconstructed images also increases rapidly with the number of low resolution images, which demonstrates the benefits of the proposed solution in super-resolution schemes.

Published in:
Int. Conf. on Pattern Recognition
Presented at:
19th International Conference on Pattern Recognition, Tampa, Florida, USA, December 8-11, 2008
IBM Best Student Paper Award

Note: The status of this file is: EPFL only

 Record created 2008-12-17, last modified 2019-03-16

Download fulltextPDF
External link:
Download fulltextURL
Rate this document:

Rate this document:
(Not yet reviewed)