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

Image defogging method and system based on deep learning neural network

A neural network and deep learning technology, applied in the field of image defogging methods and systems based on deep learning neural networks, can solve problems that cannot be directly used in the defogging process

Inactive Publication Date: 2018-03-02
CHANGCHUN INST OF OPTICS FINE MECHANICS & PHYSICS CHINESE ACAD OF SCI
View PDF3 Cites 30 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, none of these methods can be directly utilized in the dehazing process of a single image

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 and system based on deep learning neural network
  • Image defogging method and system based on deep learning neural network
  • Image defogging method and system based on deep learning neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0054] First introduce the theoretical background of the present invention: in order to describe the information of the fog map, the atmospheric scattering model can be written in the following form:

[0055] I(x)=J(x)t(x)+α(1-t(x)) (1)

[0056] t(x)=e -βd(x) (2)

[0057] Among them, I(x) is the observed fog map, J(x) is the real image that needs to be restored, t(x) is the medium transmittance, α is the global atmospheric light constant, and x represents the fog map I(x) For each pixel in , β is the atmospheric scattering coefficient, and d(x) is the depth of field. In formula (1), there are three unknown parameters. After estimating t(x) and α, the real scene graph J(x) can be recovered. Equation (2) shows that t(x) tends to 0 as d(x) tends to infinity. Comprehensive (1) (2) can be obtained.

[0058] α=I(x), d(x)→∞ (3)

[0059] In the actual process, d(x) cannot tend to infinity, but a small transmittance t will be generated at a long distance 0 . Different from the...

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 an image defogging method and a system based on a deep learning neural network. The method comprises the following steps of inputting an image with fog into a deep learning neural network system; using the deep learning neural network system to carry out characteristic extraction on the image with fog, and carrying out autonomous learning and extracting a fog correlation characteristic; carrying out multiscale mapping on the image with fog, extracting the characteristic of the image with fog in a concentrative mode under different scales and forming a characteristic graph; carrying out local extremum on each pixel in the characteristic graph, maintaining a resolution to be unchanged and acquiring the processed image; carrying out nonlinear regression operation on the processed image and acquiring initial transmissivity t(x); using a guided filter to optimize the transmissivity and carrying out image smoothing processing on the processed image; calculating an atmospheric light parameter; and according to the initial transmissivity t(x) and the atmospheric light parameter, recovering a fogless image. In the invention, connection is established between the system and an existing defogging method, and under the condition that efficiency and easy implementation are guaranteed, compared with the existing method, the method has better defogging performance.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to an image defogging method and system based on a deep learning neural network. Background technique [0002] Fog is a common atmospheric phenomenon. The ubiquity of dust, smoke or other particles in the air reduces the clarity of the atmosphere. In the field of scene photography, the transmittance of light in the atmosphere is particularly important for the imaging of distant objects, so fog usually brings many problems to imaging, such as: when the light is scattered by particles in the atmosphere, the object is in the visual field when imaging A decrease in contrast in effect. Therefore, the dehazing work has great application prospects in the field of photography and computer vision. [0003] Because the transmission degree of fog in the atmosphere is often related to the depth of field, in different positions, the greater the depth, the thicker the fog....

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
CPCG06T2207/20081G06T2207/20084G06T5/73
Inventor 任航宋玉龙郭巳秋刘博超
Owner CHANGCHUN INST OF OPTICS FINE MECHANICS & PHYSICS CHINESE ACAD OF SCI
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