Image super-resolution reconstruction method based on residual convolutional neural network

A convolutional neural network and low-resolution image technology, applied in the field of image processing and computer vision, can solve the problems of shallow network, slow network training speed, gradient disappearance, etc., and achieve the effect of accelerating convergence, avoiding loss, and easy extraction

Active Publication Date: 2022-06-28
TIANJIN UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the existing image super-resolution reconstruction algorithm based on convolutional neural network has problems such as shallow network, less available context information, image preprocessing, slow network training speed and gradient disappearance.

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 reconstruction method based on residual convolutional neural network
  • Image super-resolution reconstruction method based on residual convolutional neural network
  • Image super-resolution reconstruction method based on residual convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0029] The invention realizes an image super-resolution reconstruction method based on the residual convolutional neural network, improves the residual unit structure of the convolutional neural network, and learns the low-resolution image and the low-resolution image through the connection of multiple residual units. The mapping relationship between high-resolution images is used to reconstruct the high-resolution image using the learned mapping relationship.

[0030] 1 Image preprocessing

[0031] Before network training, the low-resolution image needs to be preprocessed. The specific process is to use the bicubic interpolation method to expand the low-resolution image to the corresponding size image. Then convert the image from RGB space to YCbCr space. Since the human eye is more sensitive to brightness information, only the Y channel of the image is processed and used as the input of the network.

[0032] 2 Feature extraction

[0033] For the input low-resolution image,...

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 belongs to the technical field of image processing and computer vision. In order to propose a new technical solution, the mapping relationship between low-resolution images and high-resolution images is learned through a multi-layer convolutional neural network, and the low-resolution images are used as the input of the network. , output high-resolution images with rich high-frequency information, and improve image reconstruction quality and visual effect. For this reason, the technical scheme adopted by the present invention is, based on the image super-resolution reconstruction method of the residual convolutional neural network, the mapping relationship between the low-resolution image and the high-resolution image is learned through the connection of multiple residual units , using the learned mappings to reconstruct high-resolution images. The invention is mainly applied to image processing occasions.

Description

technical field [0001] The invention belongs to the technical fields of image processing and computer vision, and relates to an image super-resolution reconstruction method based on a convolutional neural network. Background technique [0002] Image super-resolution reconstruction technology is the process of reconstructing high-resolution images from single-frame or multi-frame low-resolution images. Compared with low-resolution images, reconstructed images have richer high-frequency detail information, so they have a wide range of applications in computer vision and image processing. Image super-resolution reconstruction is a computer vision problem with theoretical and practical value. Image super-resolution reconstruction techniques can be divided into three categories: interpolation-based methods, reconstruction-based methods, and learning-based methods. [0003] Interpolation-based methods mainly include bilinear interpolation, bicubic interpolation, nearest neighbor...

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): G06T3/40G06N3/04G06N3/08
CPCG06T3/4046G06T3/4053G06N3/08G06N3/045
Inventor 郭继昌吴洁郭春乐朱明辉
Owner TIANJIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap