Image denoising with modified grey wolf optimizer

In todays technology, while the success of the computers for arithmetic and logical calculations, is millions or billions times higher than the humans, in the learning methods which can't be programmed with an algorithm, human outperforms computers. Some of the areas, which human outperforms computers are audio, image and odor processing. While human can detect and recognize the objects very fast, computers can't do these process with hundred percent accuracy. Some of the problems in image processing studies, is noise produced during the image acqusition and image transmission. In this study, a method based on optimization is proposed for image denosing. In most of the studies in image processing for successfully applications, optimization methods have been used. In this study, image denosing has been realized with Grey Wolf Optimizer, Modified Grey Wolf Optimizer and Genetic Algorithm. The main problem in this study is to denoise the images (noised with some Gaussian noise) with convolution with a trained filters which coefficient optimized with the optimization algorithms. Adding Gaussian noise with several test images, these images have been denoised approximating the images to the original ones with Grey Wolf Optimizer, Modified Grey Wolf Optimizer, Genetic Algorithm and Weiner Filter. The results of applications of used algorithms on the noisy images have been submitted and the results have been compared with PSNR values. The best performing algorithm has been specified. As a result, according to the attained findings, the applicability of the algorithms using for image denoising has been submitted with the PSNR values.