Research on Neural Network Based Image and Video Denoising

Abstract

Noises are inevitable in images and videos. Many denoising algorithms have been proposed. As the acquirement of image and video qualities gets higher, algorithms with high performances and easy implementation are the new trend. With the development of deep learning, neural network has been applied in denoising algorithms. These methods represent better performances and obtain video of high quality. In this paper, we will discuss several denoising methods for both images and videos on their architectures, analyze their features and comment. Finally, we will carefully forecast what neural network based denoising modules can be built in the future study.