Image Denoise Methods Based on Deep Learning

Image denoising is widely used in image, video, nuclear magnetic imaging and so on. In the application scene, camera jitter, the rapid motion of objects, dark light environment and so on may cause the captured photos to be unclean, so the research of image denoising has essential research value. This paper reviews the related research in this field in recent years, introduces the basic theory of image denoising, lists the common image noise, and then summarizes some classical denoising algorithms from traditional denoising methods. In addition, the shortcomings of traditional methods are analyzed. After that, the image denoising method based on depth learning is summarized, including the image denoising method based on REDNet, DnCNN, CBDNet, GAN, Noise2Noise structure, the principle and structure of various methods are introduced. Finally, the challenges of image denoising are analyzed, and the future research direction has prospected.