Determination of Optimum Step Size of LMS Algorithm in De-noising of Image

Abstract

If the source of noise of an image is known then we consider that the noise is correlated with one portion of the noisy image. In this case the noisy of the signal or image is removed using adaptive algorithm. In this project work we removed noise from noisy images using LMS(Least Mean Square) algorithm then similarity between original and recovered images is measured using cross-correlation efficient. The step size of the algorithm is varied to get maximum value of cross-correlation co-efficient to achieve optimum step size.