Single-image super-resolution reconstruction method based on lightweight neural network and Transform

A super-resolution reconstruction and neural network technology, applied in neural learning methods, biological neural network models, neural architectures, etc., can solve the problems of large amount of calculation and many parameters, and achieve less network parameters, less calculation amount, and spatial resolution high rate effect

Pending Publication Date: 2022-08-02
NANJING UNIV OF INFORMATION SCI & TECH
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The self-attention mechanism in the Transformer module can effectively overcome the limitations brought by the convolution inductive bias, but at the same time, there are many parameters and a large amount of calculation. Therefore, the present invention combines the lightweight network and the Transformer model, and proposes a method based on Single Image Super-resolution Reconstruction Method with Lightweight Neural Network and Transformer

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
  • Single-image super-resolution reconstruction method based on lightweight neural network and Transform
  • Single-image super-resolution reconstruction method based on lightweight neural network and Transform
  • Single-image super-resolution reconstruction method based on lightweight neural network and Transform

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0026] The solution will be described below with reference to the accompanying drawings and specific embodiments.

[0027] like figure 1 As shown, the present invention discloses a single-image super-resolution reconstruction method based on a lightweight neural network and Transformer. First, the present invention uses bicubic downsampling to obtain a low-resolution image from an original high-resolution image, and the obtained The low-resolution images are used as the input of the network, and the original high-resolution images are used as the ground-truth annotation data when the network (including the three-layer convolutional neural network, the main network, and a convolutional layer for final fusion) is trained. Secondly, the low-frequency feature extraction module (three-layer convolutional neural network) is used to extract the spatial structure features of the low-resolution image, and then the main network (multiple Mobile-T models) is used to extract the high-freq...

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 a single image super-resolution reconstruction method based on a lightweight neural network and a Transform. Firstly, a low-frequency feature extraction module is used for extracting a low-frequency feature map of a low-resolution image, secondly, the feature map is input into a frame formed by combining a depth separable convolution module and a Transform in a blocking mode, high-frequency texture detail features of the image are extracted, and finally, the low-frequency feature map and the high-frequency feature map are subjected to jump connection to output a high-resolution image. According to the method, super-resolution reconstruction of the low-resolution image can be realized, and the high-resolution image with rich details, clear texture and high spatial resolution is obtained.

Description

technical field [0001] The invention belongs to the field of image processing, and in particular relates to a single image super-resolution reconstruction method based on a lightweight neural network and a Transformer. Background technique [0002] In recent years, with the vigorous development of Internet technology, images have become one of the important sources for people to obtain information. However, due to the influence of hardware performance, environmental noise, transmission method and storage method, most images will undergo a degradation process, resulting in a reduction in the quality of the image. People often cannot obtain high-resolution images. Therefore, the reconstruction of image resolution is a A research focus and difficulty in the field of image processing. Through various software algorithms to improve the resolution of the original image at low cost, it is called Super-Resolution (SR) reconstruction. Super-resolution reconstruction technology brea...

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): G06T3/40G06N3/04G06N3/08
CPCG06T3/4053G06T3/4007G06T3/4046G06N3/08G06N3/047G06N3/048G06N3/044G06N3/045
Inventor 黄为伟郑钰辉
Owner NANJING UNIV OF INFORMATION SCI & TECH
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