Abstract

The sources of noise present substantial obstacles for image denoising. Gaussian, impulse, and speckle noise are particularly complex sources of noise in imaging. Convolutional neural networks (CNN) have gained popularity in picture denoising tasks. Several CNN denoising approaches have been investigated. These approaches evaluated using various datasets. In this research, we present an in-depth investigation of various CNN approaches for speckle noise removal in picture denoising. Different CNN image denoising algorithms were classified and analyzed. The study investigated popular datasets for testing CNN image denoising methods. A few works on CNN image denoising were chosen for evaluation and analysis. CNN approaches' motivations and concepts were described. We proposed a CNN-based picture denoising review.