Unsupervised image denoising method based on uncertainty perception

By constructing an unsupervised image denoising method of uncertainty perception, using the uncertainty graph weighted objective function, the problem of ignoring the accidental uncertainty of the training data set in the prior art is solved, and the clear texture structure of the denoising result and higher visual quality are achieved.

CN114820355BActive Publication Date: 2025-03-21FUDAN UNIVERSITY
0 Cites 0 Cited by

Patent Information

Application Number
CN202210370646.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-04-10
Publication Date
2025-03-21
Estimated Expiration
2042-04-10

AI Technical Summary

Technical Problem

The existing image denoising methods ignore the accidental uncertainty in the training dataset during the denoising process, resulting in the denoising result being too smooth and the visual effect is poor.

Method used

By constructing an unsupervised image denoising method of uncertainty perception, including extracting smooth noise modules, modeling accidental uncertainty networks and fine denoising networks, the uncertainty graph weighting objective function is used to reduce the impact of uncertainty on the denoising process.

Benefits of technology

The clear texture structure and higher visual quality of the denoising result are achieved, which reduces the interference of uncertainty on the network and makes the denoising result more reliable.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The present invention belongs to the field of digital image processing technology, and specifically is an unsupervised image denoising method based on uncertainty perception. The method of the present invention comprises: constructing a smooth noise extraction module and a rough denoising network to simulate noise blocks and perform rough denoising on images; constructing a network for modeling accidental uncertainty, estimating pixels with high uncertainty in the noise image, generating an uncertainty map, and guiding the next stage of the network to perform denoising; constructing a fine denoising network, and guiding the network to reduce the influence of uncertainty on the denoising process through the weighted objective function of the uncertainty map, so that the network can achieve better denoising effect. Experimental results show that the present invention can effectively remove noise from images, and the denoised images have good texture structure and visual quality.
Need to check novelty before this filing date? Find Prior Art