Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

An improved method of approximate information transfer algorithm based on deep learning denoising

An information transfer and deep learning technology, applied in computing, image enhancement, image analysis and other directions, can solve problems such as rate-distortion performance gap, and achieve the effect of enhancing matching degree, expanding applicability, and good denoising effect.

Active Publication Date: 2021-07-13
XI AN JIAOTONG UNIV
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, whether it is a traditional algorithm or a combination of traditional algorithm and deep learning, compared with traditional image coding technology, there is a big gap in rate-distortion performance. Therefore, how to obtain images with higher visual quality at low sampling rates is of great importance. meaning and value

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • An improved method of approximate information transfer algorithm based on deep learning denoising
  • An improved method of approximate information transfer algorithm based on deep learning denoising
  • An improved method of approximate information transfer algorithm based on deep learning denoising

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0045] The present invention provides an enhancement method based on the approximate information transfer algorithm (LDAMP) of deep learning denoising. Firstly, through the detailed stratification of the noise level, the optimized denoising model is used to fully learn the noise of each level Then replace the denoising model in the LDAMP method with the trained denoising model, and then iterate according to the iterative strategy in DAMP, finally get the reconstructed image of the measured value, and achieve the reconstruction image quality compared with LDAMP at the same sampling rate obvious improvement. The effectiveness of this scheme can also be seen from the experimental link. It provides an effective idea to solve the problem of high-quality restoration of compressed sensing images under low sampling rate under conditions in reality.

[0046] see figure 1 , the present invention a kind of enhancement method based on the approximation information transfer algorithm (LD...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an improved method of approximate information transfer algorithm based on deep learning denoising, which divides the noise into multiple levels and optimizes the structure of the denoising model so that it can better remove the noise of each level, and then trains A good denoising model is substituted into LDAMP to iteratively recover compressed sensing image measurements. In the iterative process, a more detailed noise level enhances the matching degree between the noisy image and the denoising model in each iteration, and a more complete denoising model also further improves the restoration effect of the noisy image. In the case of a low image sampling rate, the more detailed noise layering and the optimized denoising model provided by the method of the present invention provide more refined and accurate options for image preprocessing. Compared with the original LDAMP, the proposed LDAMP enhancement method of the present invention can remarkably improve the image restoration quality based on compressed sensing under the same sampling rate.

Description

technical field [0001] The invention belongs to the technical field of image compression, and in particular relates to a method for enhancing an approximate information transfer algorithm LDAMP based on deep learning denoising through detailed noise layering and denoising model optimization. Background technique [0002] Compressed sensing technology refers to the technology of reconstructing signals or images at a sampling rate lower than Nyquist. It is widely used in image processing, image retrieval, CT image reconstruction and other fields. The peak signal-to-noise ratio (PSNR) is an important indicator for judging image quality. In the field of image compression perception, the higher the PSNR of image restoration at the same sampling rate, the clearer the image restoration and the better the performance of the compression algorithm. In the process of image transmission, due to the influence and limitation of hardware, transmission bandwidth and external environment, th...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06T5/00
CPCG06T5/002G06T2207/20081G06T2207/20084
Inventor 侯兴松刘皓琰
Owner XI AN JIAOTONG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com