Implementation of Blind Deconvolution Methodology in Restoration of Blur Images

Abstract

Methods for restoring images can be classified as specialised procedures if the results are obtained in a single phase. As a result, distinguishing a clear, clean image from a noisy, blurred image is critical. Blurring is a loss in image bandwidth caused by poor image quality as a result of the inaccurate image formation process. When a digital image is captured, blurring is common. Along with these blurring effects, noise frequently corrupts every collected image. The reconstruction approach is separated into two groups: non-blind, in which the blurring feature is given and the decay process is reversed using one of the restoration algorithms, and blind, in which the blurring operator is not known. Blind system deconvolution is a highly sophisticated process that restores objects with little or no prior information of the fading PSF. The PSF is the impulse response of a point source. In this project, the Blind Deconvolution method was used to deblur a single image. PSNR and MSE costs were calculated.