A Frequency Domain Approach to Registration of Aliased Images with Application to Super-resolution
- PDF / 3,635,651 Bytes
- 14 Pages / 600.03 x 792 pts Page_size
- 48 Downloads / 249 Views
A Frequency Domain Approach to Registration of Aliased Images with Application to Super-resolution 1 and Martin Vetterli1, 2 ¨ Patrick Vandewalle,1 Sabine Susstrunk, 1 Ecole
Polytechnique F´ed´eral de Lausanne, School of Computer and Communication Sciences, 1015 Lausanne, Switzerland of Electrical Engineering and Computer Sciences, University of California, Berkeley, CA 94720-1770, USA
2 Department
Received 27 November 2004; Revised 4 May 2005; Accepted 18 May 2005 Super-resolution algorithms reconstruct a high-resolution image from a set of low-resolution images of a scene. Precise alignment of the input images is an essential part of such algorithms. If the low-resolution images are undersampled and have aliasing artifacts, the performance of standard registration algorithms decreases. We propose a frequency domain technique to precisely register a set of aliased images, based on their low-frequency, aliasing-free part. A high-resolution image is then reconstructed using cubic interpolation. Our algorithm is compared to other algorithms in simulations and practical experiments using real aliased images. Both show very good visual results and prove the attractivity of our approach in the case of aliased input images. A possible application is to digital cameras where a set of rapidly acquired images can be used to recover a higher-resolution final image. Copyright © 2006 Patrick Vandewalle et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
1.
INTRODUCTION
Image resolutionis one of the limitingparameters in digital camera design. With most digital cameras, however, it is possible to take bursts of multiple pictures in a very short period of time. Thus, high-resolution images can be reconstructed from a series of low-resolution images using super-resolution algorithms. The idea behind super-resolution imaging is to combine the information from a set of slightly different low-resolution images of the same scene and use it to construct a higherresolution image. Throughout this paper, a higher-resolution image is defined as an image with more resolving power. This means that an image that is obtained by merely upsampling and interpolating a low-resolution image does not have a higher resolution than its original. It has a larger number of pixels, but the resolving power remains the same; that is, the interpolated image does not contain more details than its original. The resolving power of an image can be increased by adding high-frequency information typically based on knowledge about the specific image model. A higher resolving power is also obtained when the aliasing ambiguity in an image is removed. We will take this second approach to construct high-resolution images. The aliasing ambiguity in an image is removed by incorporating the additional information obtained from other images of the same scene.
There are two major, and to some extent,
Data Loading...
