Ultrahigh-resolution image denoising method based on pixel recombination sampling partitioning

Through the collaborative method of pixel recombination sampling and chunking and deep learning network, ultra-high resolution images are denoised, solving the problems of computing resource limitations and non-local information utilization in the prior art, and achieving a more efficient denoising effect.

CN119671893AActive Publication Date: 2025-03-21NANJING UNIV
5 Cites 0 Cited by

Patent Information

Application Number
CN202411722224.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-11-28
Publication Date
2025-03-21
Estimated Expiration
2044-11-28

AI Technical Summary

Technical Problem

The prior art is difficult to effectively process ultra-high resolution images within limited computing resources, and the block-based denoising method cannot fully utilize non-local related information, resulting in a denoising performance.

Method used

Using a method based on pixel recombination sampling and chunking, ultra-high resolution images are recombined and downsampled and chunked, pixels with far-reaching geometric distances are aggregated in the same image block, and adjacent pixels are used to assist in denoising, and trained through a dual-channel in-1-channel deep learning network.

Benefits of technology

The local information and non-local information of ultra-high resolution images are fully utilized, which improves the denoising performance and overcomes the problem of using non-local information and traditional downsampling to destroy the image detail texture during block denoising.

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

Abstract

The invention discloses an ultrahigh-resolution image denoising method based on pixel recombination sampling and partitioning. The method comprises the following specific steps: carrying out pixel recombination down-sampling partitioning on paired noisy and noiseless ultrahigh-resolution images to generate a data set; constructing a deep learning network which uses the image blocks and the image blocks with adjacent pixels and is composed of two denoising networks and a fusion network; jointly training the whole network based on three loss functions which respectively constrain the two denoising networks and the fusion network; when reasoning is carried out, pixel recombination down-sampling blocking is carried out on a noisy ultrahigh-resolution image, image blocks and image blocks adjacent to pixels of the image blocks are sent into a network to be denoised, and pixel recombination up-sampling splicing is carried out, so that a total image denoising result of the ultrahigh-resolution image is obtained. According to the method, the local information and the non-local information of the ultrahigh-resolution image are fully utilized, and the defects that the non-local information is difficult to utilize during block denoising and the detail textures of the image are damaged by traditional downsampling are overcome.
Need to check novelty before this filing date? Find Prior Art