A Review and Taxonomy of Image Denoising Techniques

Abstract

In this paper an overview is presented on image denoising. alongside, the highlight of techniques for improving image denoising are discussed. Image denoising is a process that removes noise from noisy image. The important thing is to keep detail information about image. However, the challenges of image denoising are still not significantly improved. We present an overview on image denoising techniques. These techniques include filters such as; Median, Mean, Baysion, Guide, Gaussian, and collaborative filters with different types of noise such as; Gaussian, salt and peppers, speckle, realistic noise. There are advantages and disadvantages for each one. Also, there are techniques that have been developed to denoise image. Artificial neural network (ANN), convolution neural network (CNN), Artificial Intelligence (AI), Coccuo search (CSA), fuzzy algorithms are used to improve image denoising. Many techniques are developed to reduce noise and enhanced image quality such as median filters, wiener filter, non-local filter, and wavelet threshold-based strategies. These techniques and type of noise are discussed in this work. The parameters that relate with image denoising are discussed as well such as; Peak signal noise ratio (PSNR), Mean square error (MSE), structure similarity index measure (SSIM), Root Mean square error (RMSE). The significant techniques to improve image denoising are combination Non-Local Mean filter with convolution neural network, and convolution neural network filter with non-local filter.