Salt and Pepper Noise Removal by Combining Genetic Algorithms - Neural Networks and Statistical Methods

Abstract

This paper presents an innovative method for salt and pepper noise removal, by combining neural networks, evolutive and statistical methods. The neural network used to detect the pixels affected by noise is Pulse Coupled Neural Network, whose parameters have been optimized using a Genetic Algorithm. The pixels that are affected by noise are corrected using a Gaussian Kernel. This algorithm proves to be a very efficient method of removing salt and pepper noise from images without compromising the quality of the unaffected pixels. The method has been tested on images from the Mars rover. These tests show how the image retains fine details of the scene even if the noise levels are high.