计算机科学

计算机科学 ›› 2023, Vol. 50 ›› Issue (6A): 220600171-6.doi: 10.11896/jsjkx.220600171

• 图像处理&多媒体技术 • 上一篇    下一篇

基于对比学习的低光照图像增强

吴巨峰1,3, 赵训刚1,3, 周强1,3, 饶宁1,2   

  1. 1 桥梁结构健康与安全国家重点实验室 武汉 430050;
    2 武汉工程大学计算机科学与工程学院 武汉 430205;
    3 中铁大桥科学研究院有限公司 武汉 430034
  • 出版日期:2023-06-10 发布日期:2023-06-12
  • 通讯作者: 饶宁(1353935562@qq.com)
  • 作者简介:(33513750@qq.com)
  • 基金资助:
    湖北省科技重大专项(2020ACA006)

Contrastive Learning for Low-light Image Enhancement

WU Jufeng1,3, ZHAO Xungang1,3, ZHOU Qiang1,3, RAO Ning1,2   

  1. 1 State Key Laboratory of Bridge Structure Health and Safety,Wuhan 430050,China;
    2 School of Computer Science and Engineering,Wuhan Institute of Technology,Wuhan 430205,China;
    3 China Railway Bridge Scientific Research Institute Co.,Ltd,Wuhan 430034,China
  • Online:2023-06-10 Published:2023-06-12
  • About author:WU Jufeng,born in 1983,postgraduate,senior engineer.His main research interest is bridge information technology. RAO Ning,born in 1998,postgraduate.Her main research interest is image restoration.
  • Supported by:
    Hubei Province Science and Technology Major Project(2020ACA006).

PDF (PC)

657

摘要： 针对低照度条件下获取的图像能见度低和质量差的问题,提出了一种基于对比学习的低光照图像增强方法。文中将图像转换任务的方法运用于低光照图像增强,其挑战在于低光域和正常光域之间的差异对于像素级恢复来说过于巨大和复杂,因此所提方法将其分为两步。首先使用基于Retinex理论的传统算法对原始低光图像进行初步地光照增强,以缩小两个域之间的差异,获取低光域和正常光域之间的中间状态。然后基于对比学习将后续的增强任务分解成两个阶段,即内容增强和降质学习,以此实现两个域之间的映射。对比学习可以进一步加强网络的表征能力,最终达到高自然度的图像恢复。大量实验证明了所提方法的高效性,其能够有效地增强低光照图像,图片质量和细节保留能力优于多种先进的光照增强方法。

关键词: 低照度, 图像增强, 图像转换, 对比学习, 降质学习

Abstract: Insufficient lighting in image capturing can significantly degrade the visibility and quality of images.To tackle this problem,this paper proposes a low-light image enhancement network based on contrastive learning.It is a challenging work to apply image-to-image translation task to image enhancement since the gap between low and normal light is too huge and complex for pixel-level restoration.Therefore,the proposed method takes in two steps.In order to build intermediate states that lie between the low and normal light,this paper first adopts a traditional method based on Retinex theory to initially enhance the low-light images.Second,in order to make the mappings between two domains,the subsequent enhancement is decomposed into two stages,content enhancement and degradation learning.This work is based on contrastive learning,which can enhance the representation ability of the networks,and achieves high-naturalness recovery.Extensive experimental results demonstrate the efficiency of proposed method,which can enhance the low-light image effectively with better image quality and detail restoration ability than the SOTA low-light image enhancement methods.

Key words: Insufficient lighting, Image enhancement, Image-to-Image translation, Contrastive learning, Degradation learning

中图分类号: 

  • TP37
[1]IBRAHIM H,KONG N S P.Brightness preserving dynamic histogram equalization for image contrast enhancement[J].IEEE Transactions on Consumer Electronics,2007,53(4):1752-1758.
[2]ARICI T,DIKBAS S,ALTUNBASAK Y.A histogram modification framework and its application for image contrast enhancement[J].IEEE Transactions on Image Processing,2009,18(9):1921-1935.
[3]RAHMAN S,RAHMAN M M,ABDULLAH-AL-WADUD M,et al.An adaptive gamma correction for image enhancement[J].EURASIP Journal on Image and Video Processing,2016,2016(1):1-13.
[4]LAND E H.The retinex theory of color vision[J].ScientificAmerican,1977,237(6):108-129.
[5]GUO X,LI Y,LING H.LIME:Low-light image enhancement via illumination map estimation[J].IEEE Transactions on Image Processing,2016,26(2):982-993.
[6]WANG S,ZHENG J,HU H M,et al.Naturalness preserved enhancement algorithm for non-uniform illumination images[J].IEEE Transactions on Image Processing,2013,22(9):3538-3548.
[7]LORE K G,AKINTAYO A,SARKAR S.LLNet:A deep autoencoder approach to natural low-light image enhancement[J].Pattern Recognition,2017,61:650-662.
[8]LV F,LU F,WU J,et al.MBLLEN:Low-Light Image/Video Enhancement Using CNNs[C]//BMVC.2018:220.
[9]WEI C,WANG W,YANG W,et al.Deep retinex decomposition for low-light enhancement[C]//British Machine Vision Confe-rence.2018.
[10]WANG R,ZHANG Q,FU C W,et al.Underexposed photo enhancement using deep illumination estimation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:6849-6857.
[11]ZHANG Y,ZHANG J,GUO X.Kindling the darkness:A practical low-light image enhancer[C]//Proceedings of the 27th ACM International Conference on Multimedia.2019:1632-1640.
[12]ZHANG Y,GUO X,MA J,et al.Beyond Brightening Low-light Images[J].International Journal of Computer Vision,2021,129(4):1013-1037.
[13]RONNEBERGER O,FISCHER P,BROX T.U-net:Convolu-tional networks for biomedical image segmentation[C]//International Conference on Medical image computing and computer-assisted intervention.Cham:Springer,2015:234-241.
[14]WANG Y,CAO Y,ZHA Z J,et al.Progressive retinex:Mutually reinforced illumination-noise perception network for low-light image enhancement[C]//Proceedings of the 27th ACM International Conference on Multimedia.2019:2015-2023.
[15]FAN M,WANG W,YANG W,et al.Integrating semantic segmentation and retinex model for low-light image enhancement[C]//Proceedings of the 28th ACM International Conference on Multimedia.2020:2317-2325.
[16]RAO N,LU T,ZHOU Q,et al.Seeing in the Dark by Component-GAN[J].IEEE Signal Processing Letters,2021,28:1250-1254.
[17]JIANG Y,GONG X,LIU D,et al.Enlightengan:Deep light enhancement without paired supervision[J].IEEE Transactions on Image Processing,2021,30:2340-2349.
[18]GUO C,LI C,GUO J,et al.Zero-reference deep curve estimation for low-light image enhancement[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:1780-1789.
[19]LI C,GUO C,LOY C C.Learning to enhance low-light image via zero-reference deep curve estimation[J].arXiv:2103.00860,2021.
[20]YANG W,WANG S,FANG Y,et al.From fidelity to perceptual quality:A semi-supervised approach for low-light image enhancement[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:3063-3072.
[21]PARK T,EFROS A A,ZHANG R,et al.Contrastive learning for unpaired image-to-image translation[C]//European Confe-rence on Computer Vision.Cham:Springer,2020:319-345.
[22]HAN J,SHOEIBY M,PETERSSON L,et al.Dual Contrastive Learning for Unsupervised Image-to-Image Translation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:746-755.
[23]GUTMANN M,HYVÄRINEN A.Noise-contrastive estima-tion:A new estimation principle for unnormalized statistical models[C]//Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics.JMLR Workshop and Conference Proceedings,2010:297-304.
[24]HE K,ZHANG X,REN S,et al.Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2016:770-778.
[25]ISOLA P,ZHU J Y,ZHOU T,et al.Image-to-image translation with conditional adversarial networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2017:1125-1134.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发