A game theoretical approach for image denoising

Abstract

How to adaptively choose optimal neighborhoods is very important to pixel-domain image denoising algorithms since too many neighborhoods may cause over-smooth artifacts and too few neighborhoods may not be able to efficiently remove the noise. While the Stein's principle is shown to be able to estimate the true mean square error (MSE) for determining the optimal neighborhoods, there exists a trade-off between the accuracy of the estimate and the minimum of the true MSE. In this paper, we study the impact of this trade-off and formulate the image denoising problem as a coalition formation game. In the game, every pixel is treated as a player, who tries to seek partners to form a coalition to achieve better denoising results. By forming a coalition, every player in the coalition can obtain a gain of improving the accuracy of the Stein's estimate while incurring a cost of increasing the minimum of the true MSE. We also propose a heuristically distributed approach for coalition formation. Finally, experimental results show that the proposed game theoretical approach can achieve better performance than the nonlocal method in terms of both PSNR and visual quality.