A method and terminal for image defogging

A technology in images and images, applied in the field of image processing, can solve problems such as low calculation efficiency, overexposure of defogged images, and invalid sky areas, etc., to achieve the effects of increasing defogging speed, avoiding overexposure, and reducing color distortion

Active Publication Date: 2021-06-15
SANLI VIDEO FREQUENCY SCI & TECH SHENZHEN
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this algorithm has the following defects: first, the calculation efficiency is very low, and it cannot be applied in the industry; second, the dark channel prior knowledge is not valid for the sky area, and the sky area has serious color distortion after dehazing; third, the atmospheric light estimation method It is too simple, and the defogged image is often overexposed or too dark due to incorrect calculation of atmospheric light values
[0003] Another method is to replace the soft matting method with low computational efficiency with guided filtering, which greatly improves the computational efficiency, but for large-resolution images, the computational speed of this method is still relatively slow
[0004] Another method is to use quadtree search to calculate the atmospheric light value, which improves the accuracy of atmospheric light value calculation, but the problem of low calculation efficiency has not been solved

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
  • A method and terminal for image defogging
  • A method and terminal for image defogging
  • A method and terminal for image defogging

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0108] Please refer to figure 1 , a method for image defogging, comprising steps:

[0109] S1. Decompose the image pyramid on the image to be processed, and obtain the decomposed top image of the Gaussian pyramid;

[0110] Specifically, in this embodiment, the image to be processed is decomposed into an image pyramid, the image to be processed is divided into a multi-layer Gaussian pyramid, and the top layer image of the Gaussian pyramid after decomposing is obtained; wherein the number of layers of the Gaussian pyramid is determined by the image resolution, and generally takes 0-3 layers, the higher the number of layers, the lower the resolution of the obtained image;

[0111] S2. Determine whether the fog concentration in the non-sky area in the top image of the Gaussian pyramid is higher than the first preset threshold, if so, calculate the atmospheric light value and transmittance of the top image of the Gaussian pyramid, and judge the top image of the Gaussian pyramid W...

Embodiment 2

[0153] The difference between this embodiment and Embodiment 1 is that the calculation method of mist concentration is further limited:

[0154] Wherein, said judging whether the fog density in the non-sky area in the Gaussian pyramid top image is higher than the first preset threshold includes:

[0155] Calculate the fog concentration score_2:

[0156] ;

[0157] In the formula, x represents the pixels in the non-sky region, I represents the pixels in the region to be processed, c represents the color channel, max represents the maximum value of the pixels in the region to be processed, and min represents the pixels in the region to be processed The minimum value of , mean means taking the average value;

[0158] judging whether the fog concentration is higher than a first preset threshold;

[0159]Specifically, in practical applications, the input image to be processed may contain a fog-free image, and fog-free images do not need to be defogged, so it is necessary to ha...

Embodiment 3

[0161] Please refer to figure 2 , an image defogging terminal, including a memory, a processor, and a computer program stored on the memory and operable on the processor, and the processor implements the first or second embodiment when executing the computer program The various steps of the image dehazing method in .

[0162] In summary, a method and terminal for image defogging provided by the present invention obtains the top layer image of the Gaussian pyramid of the image to be processed; judges whether defogging is needed according to the fog concentration in the non-sky area in the top layer image of the Gaussian pyramid, and the fog concentration Defog processing is performed on images higher than the first preset threshold, which can avoid further defogging processing on fog-free images, and at the same time avoid data errors caused by similar features between the sky area and the foggy area; Calculate the atmospheric light value and transmittance, and if there is a ...

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 method and a terminal for defogging an image, which acquires the top layer image of a Gaussian pyramid of an image to be processed; judges whether defogging is needed according to the fog concentration of the top layer image of the Gaussian pyramid, and can avoid further defogging processing on the fog-free image; Calculate the atmospheric light value and transmittance for the top image of the Gaussian pyramid that needs to be defogged, and if there is a sky area, correct the transmittance of the sky area, and perform guided filtering on the transmittance, then remove the haze for the top image of the Gaussian pyramid, and set The image after dehazing is restored to the original resolution; among them, by dehazing the top image of the Gaussian pyramid and then restoring the dehazed image to the original resolution through upsampling, the image detail loss can be reduced while improving the dehazing efficiency , and correct the perspective ratio of the sky area of ​​the image, which can avoid the overexposure of the sky area after defogging, and greatly reduce the color distortion of the sky area.

Description

technical field [0001] The invention relates to the field of image processing, in particular to an image defogging method and terminal. Background technique [0002] One method of image defogging in the prior art is the dark channel prior theory proposed by observing and counting a large number of outdoor natural fog-free images. Combining the dark channel prior theory with the atmospheric light scattering model, the image can be realized to fog. However, this algorithm has the following defects: first, the calculation efficiency is very low, and it cannot be applied to the industrial world; second, the dark channel prior knowledge is not valid for the sky area, and the sky area has serious color distortion after fog removal; third, the atmospheric light estimation method It is too simple, and the defogged image is often overexposed or too dark due to incorrect calculation of atmospheric light values. [0003] Another method is to replace the soft matting method with low c...

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 Patents(China)
IPC IPC(8): G06T5/00G06T7/194G06T7/136
CPCG06T5/005G06T2207/20016G06T2207/20024G06T7/136G06T7/194
Inventor 袁小德张宇刘东剑
Owner SANLI VIDEO FREQUENCY SCI & TECH SHENZHEN
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