Image defogging method based on bright region segmentation

WANG Hai-qun; ZHAO Yan-qing; WANG Yi

doi:10.37188/CJLCD.2022-0297

您当前的位置：

首页 >

文章列表页 >

Image defogging method based on bright region segmentation

Image Processing | 更新时间：2023-05-09

- Image defogging method based on bright region segmentation
- Chinese Journal of Liquid Crystals and Displays Vol. 38, Issue 5, Pages: 636-643(2023)
- 作者机构：
  
  1.华北理工大学电气工程学院，河北唐山 063210
- 作者简介：
- 基金信息：
  
  Hebei University Science and technology research project(ZD2022114);Tangshan applied basic research project(21130212C)
- DOI：10.37188/CJLCD.2022-0297
  CLC：
扫描看全文
WANG Hai-qun, ZHAO Yan-qing, WANG Yi. Image defogging method based on bright region segmentation. [J]. Chinese Journal of Liquid Crystals and Displays 38(5):636-643(2023)
DOI：

WANG Hai-qun, ZHAO Yan-qing, WANG Yi. Image defogging method based on bright region segmentation. [J]. Chinese Journal of Liquid Crystals and Displays 38(5):636-643(2023) DOI： 10.37188/CJLCD.2022-0297.

摘要

针对暗通道先验去雾中存在的光晕和色彩失真问题，提出一种基于明亮区域分割的图像去雾算法。首先通过亮度阈值分割和区域生长将雾天图像分割为明亮区域与非明亮区域；然后用亮通道先验和超像素分别改进明亮区域和非明亮区域透射率的计算公式；再用加权融合的方法将这两个区域的透射率进行融合得到粗略的透射率，使用引导滤波对其进行优化，同时对雾天图像进行四叉树分割，取最终分割区域像素的亮度平均值为大气光值，通过大气散射模型复原去雾图像。实验结果表明，改进后去雾图像的峰值信噪比与改进前相比提高了6.5%，信息熵提高了2.1%，新增可见边之比提高了5.5%，梯度均值提高了5.3%。本文改进算法能够解决暗通道先验去雾中的问题，得到清晰且对比度高的去雾图像。

Abstract

Aiming at the problems of halo and color distortion in prior fog removal in dark channel， an image fog removal algorithm based on bright region segmentation is proposed. First， the foggy image is divided into bright areas and non-bright areas by brightness threshold segmentation and region growth. The formulas for calculating the transmittance of bright areas and non-bright areas are improved by using a bright channel prior and a super pixel， respectively. Then， the weighted fusion method is used to fuse the transmittance of these two areas to obtain a rough transmittance. Guided filtering is used to optimize them. At the same time， the foggy image is segmented by a quadtree. The average luminance of the final segmented area pixel is taken as the atmospheric light value， and the defog image is restored through the atmospheric scattering model. The experimental results show that the peak signal-to-noise ratio， the information entropy， the ratio of new visible edges and the gradient mean of the improved defog image are 6.5%， 2.1%， 5.5% and 5.3% higher than those of the original defog image. The improved algorithm can solve the problem of prior defogging in dark channel， and obtain clear and high contrast defogging images.

关键词

暗通道先验亮通道先验超像素大气散射模型

Keywords

dark channel priorbright channel priorsuper pixelatmospheric scattering model

references

李竹林，车雯雯，钱梦杭，等.一种改进的直方图均衡化图像去雾算法［J］.河南科学，2021，39（1）：1-6. doi: 10.3969/j.issn.1004-3918.2021.01.001http://dx.doi.org/10.3969/j.issn.1004-3918.2021.01.001

LI Z L， CHE W W， QIAN M H， et al. An improved image defogging algorithm based on histogram equalization ［J］. Henan Science， 2021， 39（1）： 1-6. （in Chinese）. doi: 10.3969/j.issn.1004-3918.2021.01.001http://dx.doi.org/10.3969/j.issn.1004-3918.2021.01.001

韦春苗，徐岩，李媛.基于小波变换的迭代融合去雾算法［J］.激光与光电子学进展，2021，58（20）：2010018. doi: 10.3788/lop202158.2010018http://dx.doi.org/10.3788/lop202158.2010018

WEI C M， XU Y， LI Y. Iterative fusion defogging algorithm based on wavelet transform ［J］. Progress in Laser and Optoelectronics， 2021， 58（20）： 2010018. （in Chinese）. doi: 10.3788/lop202158.2010018http://dx.doi.org/10.3788/lop202158.2010018

HE K M， SUN J， TANG X O. Single image haze removal using dark channel prior ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence， 2011， 33（12）： 2341-2353. doi: 10.1109/tpami.2010.168http://dx.doi.org/10.1109/tpami.2010.168

孙景荣，陈哲哲，谢林昌，等.基于图像天空区域分割的去雾算法［J/OL］.系统工程与电子技术，2021：1-12［2023-04-03］.http：//kns.cnki.net/kcms/detail/11.2422.TN.20220627.0914.006.htmlhttp://kns.cnki.net/kcms/detail/11.2422.TN.20220627.0914.006.html.

SUN J R， CHEN Z Z， XIE L C， et al. Defogging algorithm based on image sky region segmentation ［J］. Systems Engineering and Electronic Technology， 2021： 1-12［2023-04-03］. http：//kns.‍cnki.‍net/kcms/detail/11.2422.‍TN. 20220627.0914.006.html.http://kns.‍cnki.‍net/kcms/detail/11.2422.‍TN.20220627.0914.006.html.（in Chinese）

乔丹，张闯，朱晨雨.基于多尺度融合卷积神经网络的图像去雾算法［J］.液晶与显示，2021，36（10）：1420-1429. doi: 10.37188/CJLCD.2020-0347http://dx.doi.org/10.37188/CJLCD.2020-0347

QIAO D， ZHANG C， ZHU C Y. Image dehazing algorithm based on multi-scale concat convolutional neural network ［J］. Chinese Journal of Liquid Crystals and Displays， 2021， 36（10）： 1420-1429. （in Chinese）. doi: 10.37188/CJLCD.2020-0347http://dx.doi.org/10.37188/CJLCD.2020-0347

李玉峰，任静波，黄煜峰.基于深度学习的遥感图像去雾算法［J］.计算机应用与研究，2021，38（7）：2194-2199. doi: 10.19734/j.issn.1001-3695.2020.07.0270http://dx.doi.org/10.19734/j.issn.1001-3695.2020.07.0270

LI Y F， REN J B， HUANG Y F. Remote sensing image haze removal algorithm using deep learning ［J］. Application Research of Computers， 2021， 38（7）： 2194-2199. （in Chinese）. doi: 10.19734/j.issn.1001-3695.2020.07.0270http://dx.doi.org/10.19734/j.issn.1001-3695.2020.07.0270

WANG F， WANG C L， CHEN M L， et al. Far-field super-resolution ghost imaging with a deep neural network constraint ［J］. Light： Science & Applications， 2022， 11（1）： 1. doi: 10.1038/s41377-021-00680-whttp://dx.doi.org/10.1038/s41377-021-00680-w

张丹丹，章光，陈西江，等.改进YCbCr和区域生长的多特征融合的火焰精准识别算法［J］.激光与光电子学进展，2020，57（6）：061022. doi: 10.3788/lop57.061022http://dx.doi.org/10.3788/lop57.061022

ZHANG D D， ZHANG G， CHEN X J， et al. Flame identification algorithm based on improved multi-feature fusion of YCbCr and region growth ［J］. Laser & Optoelectronics Progress， 2020， 57（6）： 061022. （in Chinese）. doi: 10.3788/lop57.061022http://dx.doi.org/10.3788/lop57.061022

BIRATU E S， SCHWENKER F， DEBELEE T G， et al. Enhanced region growing for brain tumor MR image segmentation ［J］. Journal of Imaging， 2021， 7（2）： 22. doi: 10.3390/jimaging7020022http://dx.doi.org/10.3390/jimaging7020022

陈永，卢晨涛.基于超像素分割和暗亮通道结合的单幅图像去雾［J］.激光与光电子学进展，2020，57（16）：161023. doi: 10.3788/lop57.161023http://dx.doi.org/10.3788/lop57.161023

CHEN Y， LU C T. Single image dehazing based on superpixel segmentation combined with dark-bright channels ［J］. Laser & Optoelectronics Progress， 2020， 57（16）： 161023. （in Chinese）. doi: 10.3788/lop57.161023http://dx.doi.org/10.3788/lop57.161023

YAN Y Y， REN W Q， GUO Y F， et al. Image deblurring via extreme channels prior ［C］//2017 IEEE Conference on Computer Vision and Pattern Recognition （CVPR）. Honolulu， HI， USA： IEEE， 2017. doi: 10.1109/cvpr.2017.738http://dx.doi.org/10.1109/cvpr.2017.738

刘安祺，刘华勇，王焕宝.改进的SLIC超像素图像分割与合并算法［J］.安徽建筑大学学报，2020，28（4）：39-46.

LIU A Q， LIU H Y， WANG H B. Improved super-pixel image segmentation and merging algorithm ［J］. Journal of Anhui Jianzhu University， 2020， 28（4）： 39-46. （in Chinese）

LIN J C， YANG J J， YU T， et al. Image dehazing algorithm based on improved guided filtering ［J］. IOP Conference Series： Earth and Environmental Science， 2020， 571： 012024. doi: 10.1088/1755-1315/571/1/012024http://dx.doi.org/10.1088/1755-1315/571/1/012024

李鹏飞，何小海，卿粼波，等.暗通道融合亮通道优化的夜间图像去雾算法［J］.液晶与显示，2021，36（4）：596-604. doi: 10.37188/CJLCD.2020-0208http://dx.doi.org/10.37188/CJLCD.2020-0208

LI P F， HE X H， QING L B， et al. Nighttime dehazing algorithm of dark channel and bright channel fusion optimization ［J］. Chinese Journal of Liquid Crystals and Displays， 2021， 36（4）： 596-604. （in Chinese）. doi: 10.37188/CJLCD.2020-0208http://dx.doi.org/10.37188/CJLCD.2020-0208

彭静，薛奉金，苑玉彬.基于多尺度Retinex和暗通道的自适应图像去雾算法［J］.激光与光电子学进展，2021，58（4）：0410004. doi: 10.3788/lop202158.0410004http://dx.doi.org/10.3788/lop202158.0410004

PENG J， XUE F J， YUAN Y B. Adaptive image defogging algorithm combining multi-scale Retinex and dark channel ［J］. Progress in Laser and Optoelectronics， 2021， 58（4）： 0410004. （in Chinese）. doi: 10.3788/lop202158.0410004http://dx.doi.org/10.3788/lop202158.0410004

束慧铃.自然图像去雾算法及清晰化客观评价研究［D］.南京：南京邮电大学，2020. doi: 10.3934/mbe.2021053http://dx.doi.org/10.3934/mbe.2021053

SHU H L. Research on natural image dehazing algorithm and clarification objective evaluation ［D］. Nanjing： Nanjing University of Posts and Telecommunications， 2020. （in Chinese）. doi: 10.3934/mbe.2021053http://dx.doi.org/10.3934/mbe.2021053

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

A dehazing stray light algorithm for fundus retinal image

Image defogging algorithm based on improved dark channel and adaptive tolerance

Aerial image defogging based on dual-threshold position and transmittance constraint

Image defogging algorithms based on multiscale convolution neural network

Related Author

No data

Related Institution

College of Electrical Engineering， North China Univ. of Tech

Jinan Guoke Medical Technology Development Co.， Ltd

Shenyang Guoke Guangming Medical Technology Co.， Ltd

The Second Affiliated Hospital of Soochow University

Jiangsu Key Laboratory of Medical Optics， Suzhou Institute of Biomedical Engineering and Technology， Chinese Academy of Science

Address：No.3888 Dong Nanhu Road, Changchun, Jilin, China 130033 Postal code：130033
Tel：0431-86176059 Email：yjxs@ciomp.ac.cn
Technical support is provided by Beijing Founder electronics co., LTD 吉ICP备11002662号-17 京公网安备11010802024621
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰