IMAGE RESTORATION USING NON-BLIND DECONVOLUTION APPROACH – A COMPARISION

The Image restoration is the process of recovery of an image that has been corrupted by some degradation phenomenon. Degradation occurs due to motion blur, gaussian blur, noise and camera mismatch. In this paper an attempt has been made to recover image from the corrupted image using Particle Swarm Optimization (PSO) algorithm in the presence of gaussian blur. For this purpose, a heuristic particle swarm optimization technique has been developed to optimize the parameters of the Point Spread Function (PSF). Higher resolution, better quality image is obtained by deblurring the noisy/ blurred image using this method. The algorithm performance is compared with Lucy Richardson algorithm. Experimental results indicated that the PSO regularized technique will improve the image quality significantly. Better results in terms of PSNR, SNR and image quality index are achieved..