Multichannel image restoration based on optimization of the structural similarity index
IEEE Asilomar Conference on Signals, Systems and Computers, IEEE: 812--816 , 2009
Abstract: In this paper a framework for multichannel image restoration based on optimization of the structural similarity (SSIM) index is presented. The SSIM index describes the similarity of images more appropriately for the human visual system than the mean square error (MSE). It has not yet been explored for the multi channel restoration task. The construction of an optimization algorithm is difficult due to the non-linearity of the SSIM measure. The existing solution based on a quasi-convex problem formulation is successfully extended for the multichannel image restoration. The correctness of the algorithm is verified on sample images and it is shown that multi-view information can significantly improve the restoration results.
Publisher's link
Other associated files :
![Paper_multiview_ssim.pdf [196KB]](/Publications/images/pdf.png "Paper_multiview_ssim.pdf [196KB]")
Images and movies
![a_drosophila_result_noise_zero.png [24KB]](/Publications/2009/BT09/a_drosophila_result_noise_zero.png "a_drosophila_result_noise_zero.png [24KB]")
![blockdiagram.png [12KB]](/Publications/2009/BT09/blockdiagram.png "blockdiagram.png [12KB]")
BibTex reference
@InProceedings{BT09,
author = "M. Temerinac-Ott and H. Burkhardt",
title = "Multichannel image restoration based on optimization of the structural similarity index",
booktitle = "IEEE Asilomar Conference on Signals, Systems and Computers",
pages = "812--816 ",
year = "2009",
publisher = "IEEE",
organization = "IEEE",
keywords = "inverse filter , multichannel image restoration , quasi-convex optimization of non-linear functions , structural similarity ",
url = "http://lmbweb.informatik.uni-freiburg.de/Publications/2009/BT09"
}