Evaluating performance of some common filtering techniques for removal of Gaussian noise in images

Abstract

In this eork, Noise is modelled as Additive White Gaussian Noise (AWGN), where all the image pixels deviate from their original values following the Gaussian Curve. Many Gaussian noise removal techniques require the knowledge of standard deviation as a measure of noise corruption for the purpose of setting threshold value, size of the sliding window etc. We have used different filtering techniques, viz., Mean, Median, Fuzzy, Wiener and Sigma to produce a noise-free image. We found that the Weiner Filter does the best job at denoising the image from AWGN.