Rain removal system based on dual-channel multi-scale discrimination model and control method

A discriminant model and multi-scale technology, applied in the field of computer vision, can solve problems affecting the visual quality of images, image degradation, etc., and achieve the effects of improving rain removal efficiency, improving feature reuse, and reducing model parameters

Inactive Publication Date: 2021-04-13
NANCHANG HANGKONG UNIVERSITY
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem in the prior art that the image captured by the camera in rainy days is severely degraded and seriously affects the visual quality of the image, the embodiment of the present invention provides a rain removal system and control method based on a dual-channel multi-scale discriminant model

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
  • Rain removal system based on dual-channel multi-scale discrimination model and control method
  • Rain removal system based on dual-channel multi-scale discrimination model and control method
  • Rain removal system based on dual-channel multi-scale discrimination model and control method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0031]In order to make the objects, technical solutions, and advantages of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. The depth learning in the article refers to the artificial neural network composed of many layers, and "deep" refers to the number of hidden layers in the neural network. Use the neural network to simulate the human brain to process the data.

[0032]Batch normalization (BN) is for deep neural networks, with network training, the adjustment of the front layer parameters makes the distribution of the latter input data change, and the layers need to be continuous in the process of training. Change to adapt to learning this new data distribution. Batch normalization is used to reduce the changes in internal data, which can greatly speed up the training of depth neural networks, reduce gradient dependence on parameter scale or initial value, and solve the loss of lear...

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 rain removal system based on a dual-channel multi-scale discrimination model and a control method, the system comprises a generator and a dual-channel multi-scale discriminator, the generator is used for generating an adversarial network, the generator is composed of a multi-scale feature extraction unit and a dense connection unit; the multi-scale feature extraction unit is used for acquiring a receiving field and acquiring multi-scale rain mark information according to scale input; the dense connection unit is used for removing a rain layer and reserving a background layer; and the dual-channel multi-scale discriminator is provided with a dual-channel discrimination module, and the dual-channel discrimination module is used for judging whether the output rain-removed image and the real rain-free image are true or false. The influence of raindrops on images is solved by utilizing a dual-discrimination model generative adversarial network, the rain removal effect is enhanced, and the problems that images shot by a camera in rainy days are seriously degraded and the visual quality of the images is seriously influenced in the prior art are solved.

Description

Technical field[0001]The present invention relates to the field of computer visual, and more particularly to a dual-channel multi-scale discriminant model to rain system and control method.Background technique[0002]With the rapid development of science and the popularity of smartphones, mobile phones and cameras are severely degraded in rainy days, seriously degraded, seriously affecting image quality. In order to improve the overall quality of these degraded images, it is important to remove these undesired raindrops that have been enhanced and ensuring that the performance of the visual algorithm is enhanced. It is a very important issue that the raindrops that do not want to appear. .[0003]At present, the existing way to rain is divided into two categories: one is based on the rain, such as a high-frequency approach, assuming that the rain strip belongs to the high frequency component of the input image, and then learn from sparse coding and dictionary It is divided into rain and...

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): G06T5/50G06K9/46G06K9/62G06N3/04G06N3/08
CPCG06T5/50G06N3/08G06T2207/10004G06T2207/20224G06V10/44G06N3/048G06N3/045G06F18/24
Inventor 盖杉谢强强
Owner NANCHANG HANGKONG UNIVERSITY
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