A Fractional-Order Derivative Variational Framework for Blind Image Deconvolution via Nash Game Strategy

This paper introduces a novel hybrid model that addresses the image blind deconvolution problem within the framework of Nash games. Our proposed model formulates the deconvolution task as a Nash game involving two players: one player focuses on image deblurring, while the other estimates the point spread function (PSF) or blur kernel. The optimal solution to our problem is defined as a Nash equilibrium, which is achieved through an alternating minimization algorithm. Furthermore, we enhance our approach by integrating fractional-order derivatives, with the goal of improving the accuracy and robustness of image restoration. Our numerical results demonstrate the superiority and effectiveness of our methodology, yielding visually appealing outcomes and a higher Peak Signal-to-Noise Ratio (PSNR). This advancement holds promise for the future of blind image deconvolution, expanding its application and effectiveness across various fields.