Fast image defogging method based on dark channel prior and Retinex theory

A dark channel color prior, image technology, applied in the fields of navigation control, intelligent vehicles, urban traffic, video surveillance, target tracking and terrain survey, can solve the problems of color distortion, high time and space complexity, etc., to achieve good image color , the effect of improving the degree of automation and computing speed, broad prospects and application value

Inactive Publication Date: 2015-08-26
WUHAN UNIV OF TECH
View PDF4 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Image defogging based on the dark primary color prior reveals the essential characteristics of the image and has a good defogging effect, but for bright areas such as the sky and white clouds where there is no dark primary color, the restoration result will have obvious color distortion, and The soft matting method adopted has high time and space complexity

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
  • Fast image defogging method based on dark channel prior and Retinex theory
  • Fast image defogging method based on dark channel prior and Retinex theory
  • Fast image defogging method based on dark channel prior and Retinex theory

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0025] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0026] A fast image defogging method based on dark channel prior and Retinex theory ( figure 1 ), this embodiment is for foggy images ( image 3 ) is processed as follows, specifically comprising the following steps:

[0027] Step S1, obtain a simplified atmospheric scattering model, including the following steps:

[0028] Step S11, calculate the estimated value of atmospheric light, the specific method is as follows: firstly use the prior knowledge of the dark channel to determine the value range of atmospheric light through the gray scale open operation, and then calculate the atmospheric light within the value range using the method of interval estimation, That is, the estimated value of atmospheric light is obtained.

[0029] Estimation of atmospheric light value A:

[0030]

[0031] α = m e ...

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 fast image defogging method based on dark channel prior and a Retinex theory. The method comprises the following steps of: firstly obtaining a simplified atmospheric scattering model; then obtaining an initial estimation value and a rough estimation value of a medium transmission rate; carrying out pixel-stage fusion to obtain a final estimation value of the medium transmission rate; and finally substituting the final estimation value of the medium transmission rate into the simplified atmospheric scattering model for image restoration to obtain a defogged image. The method provided by the invention has the advantages that a higher operation speed is realized; the definition and the contrast ratio of a restored image can be effectively improved; meanwhile, better image colors are obtained; and wide prospects and great application values are realized in the fields of urban traffic, video monitoring, intelligent vehicles and the like.

Description

technical field [0001] The invention relates to the fields of urban traffic, video monitoring, intelligent vehicles, navigation control, target tracking and terrain survey, and in particular relates to a method for quickly defogging a single image. Background technique [0002] Smog is a common natural phenomenon, which is produced by the action of a large number of tiny water droplets, aerosols and other particles suspended in the atmosphere. Due to the absorption, scattering and refraction of light by suspended particles, the contrast, saturation and hue shift of the target in the image are reduced, which directly limits and affects the effectiveness of various outdoor vision systems. The sharpening processing of degraded images in foggy weather is of great significance to improve the reliability and robustness of outdoor vision systems, and single image dehazing has become a research hotspot in the field of image dehazing. Various methods have been proposed for image def...

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
Inventor 杨杰刘海波吴正平邓勇
Owner WUHAN UNIV OF 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