Deep network video denoising method based on model constraint and program product

Abstract

The invention belongs to a video denoising method, and provides a deep network video denoising method based on model constraint and a program product in order to solve the technical problems that the denoising effect of an existing video denoising method based on deep learning does not completely exceed that of a traditional method based on a model, and the quality of a denoised video cannot be effectively guaranteed. For video denoising, a new video denoising network based on model guidance is provided as an observation model, the video denoising network describes the relationship between adjacent noise frames, a clean video frame can be effectively recovered through maximum posterior probability estimation, and effective calculation can also be performed through iteration steps. Compared with a traditional observation model, the new observation model provided by the invention has the advantages that the calculation complexity is greatly reduced by realizing effective iteration steps, and the video quality after noise reduction can be effectively improved.

Description

technical field [0001] The invention belongs to a video denoising method, in particular to a deep network video denoising method and program product based on model constraints. Background technique [0002] Despite the ever-improving performance of imaging sensors, digital imaging still inevitably introduces noise, which not only degrades image and video quality, but also adversely affects subsequent image processing tasks. [0003] In order to improve the visual effect, researchers at home and abroad have proposed many image and video denoising algorithms. For single image denoising, existing methods are mainly divided into model-based methods and deep learning-based methods. Model-based denoising methods are highly flexible, but have high computational complexity and are difficult to jointly optimize the parameters in the algorithm. With the rise of deep learning, Convolution Neural Network (CNN) has also been used for image denoising. By learning an end-to-end mapping f...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the technical problem that the current video denoising method based on deep learning does not completely exceed the traditional model-based method, and the video quality after denoising cannot be effectively guaranteed, the present invention provides a model-based constraint Deep network video denoising method and program product

Images