Research Area: Machine Learning
The recent advances in hardware and imaging systems made the digital cameras ubiquitous. Although the development of hardware has steadily improved the quality of images for the last several decades, image degradation is unavoidable due to the many factors affecting the image acquisition process and the subsequent post-processing. Image denoising, which aims to reconstruct a high quality image from its degraded observation, is a classical yet still very active topic in the area of low-level computer vision. It represents an important building block in real applications such as digital photography, medical image analysis, remote sensing, surveillance and digital entertainment. Also, image denoising constitutes an ideal test bed for evaluating image prior modeling methods. In this paper, we briefly review recent progresses in image denoising. We firstly present an overview of prior modeling approaches used in image denoising task. Then, we review conventional sparse representation based denoising algorithms, low-rank based denoising algorithms and recently proposed deep neural networks based approaches. At last, we discuss some emerging topics and open problems about image denoising.
Keywords:
Image Denoising
Deep neural networks
Deep Learning
Author(s) Name: Shuhang Gu & Radu Timofte
Journal name:
Conferrence name: Inpainting and Denoising Challenges
Publisher name: Springer
DOI: 10.1007/978-3-030-25614-2_1
Volume Information:
