Effective hybrid genetic algorithm for removing salt and pepper noise

This study presents a new approach for recovering an image perturbed by salt and pepper noise (SPN) using a hybrid genetic algorithm (HGA) at all densities, called effective HGA (EHGA). The main contribution of the proposed algorithm is combining the genetic algorithm with image denoising methods that are integrated into the population to achieve rapid convergence. The idea is to evolve a group of individuals into a number of iterations using crossover and mutation operators. This approach evolves a set of images rather than a set of parameters from the filters. Experimental results of simulation on different images using peak signal-to-noise ratio, structural similarity index metric, image enhancement factor and Universal Quality Index show that the proposed algorithm outperforms other methods in removing the SPN qualitatively and quantitatively if the noise density is moderate and high. EHGA also preserves important features such as texture and corners of the image.