Abstract: Image processing plays a vital role in the field of biomedical, science, engineering, defence etc. Digital image becomes omnipresent and this desirable change made possible by science. Various steps include image acquisition, image restoration, image enhancement and image segmentation etc. An image signal gets corrupted with noise during acquisition, communication, storage and retrieval processes. It includes various noises like salt & pepper, Gaussian & speckle noise. To remove the noise in digital images we use various filters like mean filter, median filter, LMS adaptive filter and also various techniques. But these techniques can not improve the image quality. So that we use the sparse representation in this we can minimize the difference between the sparse codes of degraded image and the sparse code of unknown original image so that we can improve the performance of sparsity based image restoration. In this we use parameters PSNR, SSIM and noise sigma.
Keywords: Image acquisition, image restoration, image enhancement and image segmentation, parameters PSNR, SSIM and noise sigma.