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

Image super-resolution and deblurring parallel realization method

A super-resolution and deblurring technology, applied in image analysis, image enhancement, image data processing, etc., can solve the problems of missing high-frequency details of images and poor effect, and achieves reduction of computational complexity, clear output, and sharpening of feature maps. Effect

Active Publication Date: 2018-09-28
XI AN JIAOTONG UNIV
View PDF2 Cites 20 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] It is difficult for the existing image deblurring algorithm to estimate the blur kernel suitable for the whole image. On the other hand, the existing image super-resolution method will lose the high-frequency details of the image. When trying to combine these two tasks, the effect will be better. Difference

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
  • Image super-resolution and deblurring parallel realization method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment approach

[0027] According to one embodiment of the present invention, it mainly includes the following steps:

[0028] Step 1: Image data set preprocessing, such as DIV 2K high-definition image data set, after randomly cropping image blocks on each training set image and performing random flipping, as the training true value of the super-resolution branch of the neural network, the cropped The obtained image block is interpolated and scaled as the training truth value of the deblurring branch of the neural network, and then the Gaussian kernel is artificially added to it for image blurring, and Gaussian noise at a level of 0.1 is added, and the final image is used as the input of the neural network.

[0029] Step 2: Construct a neural network, use a neural network encoding-decoding module based on deep learning to extract the features of the entire input image, and invent a two-branch structure to achieve image deblurring and super-resolution tasks respectively.

[0030] Among them, th...

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 image super-resolution and deblurring parallel realization method. After a proper data set is obtained, input image features are quickly extracted through a particularly designed coding-decoding neural network module with feature information bridging, and an output feature graph serves as image super-resolution and deblurring branches at the same time to perform relatedtask processing, so that the calculation amount is reduced; and meanwhile, during network training, two branch networks proposed in the method can excite the shared feature graph in different degrees,so that the effects of the image super-resolution and deblurring branches are improved.

Description

technical field [0001] The invention belongs to the field of computer vision and image processing, in particular to a method for realizing image super-resolution and deblurring in parallel. Background technique [0002] In recent years, with the development and maturity of deep learning technology, image super-resolution and image deblurring algorithm research has received more and more attention, and great progress has been made in the algorithm. [0003] The purpose of image super-resolution is to restore high-resolution images from low-resolution images. It can not only generate satisfactory high-resolution images, but also provide deeper image processing processes such as target detection and face recognition. Provides a higher quality image source. However, long-term exploration has found that camera shake, out-of-focus, turbulence and other phenomena seriously hinder the research of image super-resolution methods. [0004] Image deblurring is a method to restore a cl...

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 Applications(China)
IPC IPC(8): G06T5/00
CPCG06T2207/20084G06T2207/20081G06T5/73
Inventor 王飞张康龙张昕昳韦昭谷宇祝捷董航
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