Filtering The Corrupted Image By Different Rate Of Gaussian Noise

Abstract

There are many reports that deal with removing noises from degraded images. The common noises reduction techniques are based on the assumption that the noise corrupts the image is signal independent, and adopte local statistics, computed in fixed neighborhood to recover the image signal.

In this paper we have generating Gaussian noise (and) with different rates (20, 40, 60, 80,100) % and we study additive Lee's filters for reducing additive Gaussian noise with a square and circular shape window. The resulting images and fidelity criteria showed that the circular smoothing window exhibit the best results in image noise reduction.