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

Image defogging method integrating image restoration and image enhancement and convolutional network thereof

A technology for network enhancement and image enhancement, applied in image enhancement, image analysis, image data processing, etc., can solve problems such as poor performance, poor results, and lack of information guidance for enhancement methods.

Pending Publication Date: 2021-01-12
ZHEJIANG GONGSHANG UNIVERSITY
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Physical model-based restoration methods require the input hazy image to conform to the physical model of the current method, otherwise it will lead to poor results
End-to-end based augmentation methods lack information guidance and may perform poorly in complex scenes

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 defogging method integrating image restoration and image enhancement and convolutional network thereof
  • Image defogging method integrating image restoration and image enhancement and convolutional network thereof
  • Image defogging method integrating image restoration and image enhancement and convolutional network thereof

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0071] In order to make the technical means, creative features, goals and effects achieved by the present invention easy to understand, the present invention will be further elaborated below in combination with diagrams and specific implementation processes.

[0072] This is an image dehazing convolutional network that combines image restoration and image enhancement. This method uses three different convolutional neural networks to achieve dehazing. The method simulates the mapping from the original foggy image to the fog-free image. Realize defogging.

[0073] In this method, the atmospheric scattering model is used as the physical model, and the atmospheric scattering model is expressed as follows:

[0074] I=J×t+A×(1-t)

[0075] Among them, I is a foggy image; J is a non-foggy image, which is the same size as the foggy image I; A is an atmospheric light image, which is the same size as the foggy image I; t is a transmission image, indicating the degree to which the foggy ...

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 relates to the technical field of image enhancement and computer vision, in particular to an image defogging method integrating image restoration and image enhancement and a convolutional network of the image defogging method. According to the method, three different convolutional neural networks are utilized to perform fusion defogging, and the method simulates the mapping from an original foggy image to a fogless image to realize defogging. According to the invention, the defogging capability of the enhanced network learning non-physical model is improved by combining the defogging graph of the restored network in the loss calculation of the enhanced network; the gating fusion network calculates self-adaptive weight graphs of the two defogging graphs of the restoration network and the enhancement network, and the weight graphs can play a role in fusing good parts of the restoration network and the enhancement network and improving defogging performance. According to themethod, the design defects existing in a physical model and an end-to-end convolutional neural defogging network at present are mainly overcome, a defogging graph is optimized, the application rangeof a defogging algorithm is expanded, and the robustness of the defogging algorithm is enhanced.

Description

technical field [0001] The invention relates to the fields of image enhancement technology and computer vision technology, in particular to an image defogging method and a convolutional network combining image restoration and image enhancement. Background technique [0002] Small particles of solids and small droplets of liquids suspended in the air cause the smog effect. The reflected light of objects will be attenuated and scattered due to haze weather, thereby reducing the visibility and contrast of the environment. When the sensor is used to collect images in hazy weather, images containing haze will be obtained, which are called foggy images. Dehazing algorithm is an image enhancement algorithm whose goal is to improve image quality, clarity, contrast, etc. without affecting image information, or as a preprocessing step for other advanced visual tasks. The input of the dehazing algorithm is a single hazy image, and its target output is a clean and clear dehazed image....

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): G06T5/00G06T5/50G06N3/04G06N3/08
CPCG06T5/50G06N3/084G06T2207/20081G06T2207/20084G06T2207/20221G06N3/045G06T5/77G06T5/73Y02A90/10
Inventor 刘春晓章理登李彪
Owner ZHEJIANG GONGSHANG UNIVERSITY
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