基于改进CycleGAN的多失真类型水下图像增强

doi:10.3785/j.issn.1008-973X.2025.06.006

浙江大学学报(工学版)

2025, Vol. 59

Issue (6): 1148-1158 DOI: 10.3785/j.issn.1008-973X.2025.06.006

计算机技术

基于改进CycleGAN的多失真类型水下图像增强

吕振鸣1(

),董绍江1,*(

),夏宗佑2,牟小燕3,王明权4

1. 重庆交通大学机电与车辆工程学院，重庆 400074
2. 重庆交通大学交通运输学院，重庆 400074
3. 重庆工业职业技术学院机械工程学院，重庆 402260
4. 重庆市勘测院，重庆 400020

Multi-distortion type underwater image enhancement based on improved CycleGAN

Zhenming LV1(

),Shaojiang DONG1,*(

),Zongyou XIA2,Xiaoyan MOU3,Mingquan WANG4

1. School of Mechantronics and Vehicle Engineering, Chongqing Jiaotong University, Chongqing 400074, China
2. College of Traffic and Transportation, Chongqing Jiaotong University, Chongqing 400074, China
3. School of Mechanical Engineering, Chongqing Vocational and Technical College of Industry, Chongqing 402260, China
4. Chongqing Survey Institute, Chongqing 400020, China

全文: PDF(4587 KB) HTML

摘要：

针对由散射、吸收和色偏等多种因素导致的水下图像模糊、对比度低、图像失真辨识困难的问题，提出基于改进CycleGAN的多失真类型水下图像增强算法. 为了提高图像增强效果，在CycleGAN的生成器中采用Auto-Encoder+Skip-connection的网络结构，添加全局色彩校正结构，从像素方面以及颜色方面进行全局增强，从而更好地捕捉水下图像中的色彩信息. 设计多维感知判别器来学习图像的全局特征和局部特征，此设计更加注重图像局部细节部分，有效针对散斑和色彩噪声，从多维空间上感知图像，提取特征能力更强，从而能提高图像判别的精度. 在EUVP、UIEB和U45数据集上的实验结果表明，所提出的方法取得了较好的结果，相较其他算法，在处理多种水下失真类型的图像方面，该算法的SSIM指标平均比第2名高出1.57%、PSNR指标高出1.836%、UIQM指标高出1.324%、UCIQE指标高出1.086%，在处理颜色和噪声细节方面表现出色.

关键词： 图像增强; CycleGAN; 全局色彩校正结构; 多维感知判别器; 多种失真

Abstract:

A multi-distortion type underwater image enhancement algorithm based on improved CycleGAN was proposed, aiming at the difficulties of underwater image blurring, low contrast and image distortion recognition caused by various factors such as scattering, absorption and color deviation. Firstly, in order to improve the image enhancement effect, Auto-Encoder+Skip-connection network structure was used in the generator of CycleGAN, and global color correction structure was added for global enhancement in terms of pixel as well as color, so as to better capture the color information in underwater images. Secondly, a multidimensional perceptual discriminator was designed to learn the global and local features of the image. This discriminator payed more attention to the local details of the image, effectively targeted scattering and color noise, perceived the image from a multidimensional space, and had a stronger ability to extract the features, thereby enhancing the accuracy of image discrimination. Finally, the experimental results on EUVP, UIEB and U45 datasets showed that the proposed method achieved better results, compared with other algorithms. In processing multi-distortion types of underwater images, the algorithm’s SSIM indicator was higher than that of the second place by an average of 1.57%, the PSNR indicator was higher by 1.836%, the UIQM indicator was higher by 1.324%, and the UCIQE indicator was higher by 1.086%. The proposed method performed well in processing color and noise details.

Key words: image enhancement CycleGAN global color correction structure multidimensional perceptual discriminator multiple distortion

收稿日期: 2024-04-08 出版日期: 2025-05-30

CLC:

TP 391.41

基金资助: 重庆市教委科学技术研究资助项目（KJQN202203207)；重庆市科技创新领军人才支持计划资助项目(CSTCCCXLJRC201920)；重庆市高校创新研究群体(CXQT20019).

通讯作者: 董绍江 E-mail: 13946501711@163.com;dongshaojiang100@163.com

作者简介: 吕振鸣（2000—），男，硕士生，从事水下机器人研究. orcid.org/0009-0009-3520-7523. E-mail：13946501711@163.com

	服务
	把本文推荐给朋友
	加入引用管理器
	E-mail Alert
	作者相关文章
	吕振鸣
	董绍江
	夏宗佑
	牟小燕
	王明权

引用本文:

吕振鸣,董绍江,夏宗佑,牟小燕,王明权. 基于改进CycleGAN的多失真类型水下图像增强[J]. 浙江大学学报(工学版), 2025, 59(6): 1148-1158.

Zhenming LV,Shaojiang DONG,Zongyou XIA,Xiaoyan MOU,Mingquan WANG. Multi-distortion type underwater image enhancement based on improved CycleGAN. Journal of ZheJiang University (Engineering Science), 2025, 59(6): 1148-1158.

链接本文:

https://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2025.06.006 或 https://www.zjujournals.com/eng/CN/Y2025/V59/I6/1148

图 1 面向多失真类型水下图像的网络总体结构

图 2 水下图像增强网络生成器模型结构

图 3 挤压和注意模块

图 4 生成器颈部的自注意力模块

图 5 水下图像增强网络判别器模型结构

图 6 黄色、有雾、绿色水下图像增强方法的视觉比较（局部差异放大表示）

图 7 蓝色、彩色水下图像增强方法的视觉比较（局部差异放大表示）

图 8 各种水下图像增强方法在U45数据集上的视觉比较（局部差异放大表示）

表 1 几种算法在EUVP和UIEB中黄色失真图像上的定量评估结果

表 2 几种算法在EUVP和UIEB中有雾失真图像上的定量评估结果

表 3 几种算法在EUVP和UIEB中蓝色失真图像上的定量评估结果

表 4 几种算法在EUVP和UIEB中彩色失真图像上的定量评估结果

表 5 几种算法在EUVP和UIEB中绿色失真图像上的定量评估结果

表 6 几种算法在U45数据集上的定量评估结果

图 9 消融实验的视觉比较（局部差异放大表示）

表 7 消融实验的定量评估

表 8 损失函数消融实验的定量评估

1	WHITCOMB L, YOERGER D R, SINGH H, et al. Advances in underwater robot vehicles for deep ocean exploration: navigation, control, and survey operations [C]// Robotics Research: the 9th International Symposium. London: Springer, 2000: 439−448.
2	LI X, HOU G, TAN L, et al A hybrid framework for underwater image enhancement[J]. IEEE Access, 2020, 8: 197448- 197462 doi: 10.1109/ACCESS.2020.3034275
3	FU X, ZHUANG P, HUANG Y, et al. A retinex-based enhancing approach for single underwater image [C]// IEEE International Conference on Image Processing. Paris: IEEE, 2014: 4572–4576.
4	PAN P, YUAN F, CHENG E Underwater image de-scattering and enhancing using dehazenet and HWD[J]. Journal of Marine Science and Technology, 2018, 26 (4): 6
5	SHEN L, REDA M, ZHANG X, et al Polarization-driven solution for mitigating scattering and uneven illumination in underwater imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 4202615
6	MENG H, YAN Y, CAI C, et al A hybrid algorithm for underwater image restoration based on color correction and image sharpening[J]. Multimedia Systems, 2022, 28 (6): 1975- 1985 doi: 10.1007/s00530-020-00693-2
7	PEREZ J, ATTANASIO A C, NECHYPORENKO N, et al. A deep learning approach for underwater image enhancement [C]// Biomedical Applications Based on Natural and Artificial Computing. Cham: Springer International Publishing, 2017: 183–192.
8	WEI D, XIE H, ZHANG Z, et al Learning a Holistic-Specific color transformer with Couple Contrastive constraints for underwater image enhancement and beyond[J]. Journal of Visual Communication and Image Representation, 2024, 98: 104059 doi: 10.1016/j.jvcir.2024.104059
9	ZHANG T, LIU Y MTUW-GAN: a multi-teacher knowledge distillation generative adversarial network for underwater image enhancement[J]. Applied Sciences, 2024, 14 (2): 529 doi: 10.3390/app14020529
10	ANWAR S, LI C, PORIKLI F, et al. Deep underwater image enhancement [EB/OL]. [2024-03-02]. https://arxiv.org/abs/1807.03528v1.
11	LIANG D, CHU J, CUI Y, et al NPT-UL: an underwater image enhancement framework based on nonphysical transformation and unsupervised learning[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 5608819
12	LIU Q, ZHANG Q, LIU W, et al WSDS-GAN: a weak-strong dual supervised learning method for underwater image enhancement[J]. Pattern Recognition, 2023, 143: 109774 doi: 10.1016/j.patcog.2023.109774
13	PARK J, HAN D K, KO H Adaptive weighted multi-discriminator CycleGAN for underwater image enhancement[J]. Journal of Marine Science and Engineering, 2019, 7 (7): 200 doi: 10.3390/jmse7070200
14	WANG H, ZHANG W, BAI L, et al Metalantis: a comprehensive underwater image enhancement framework[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 5618319
15	ZHU X, LIN M, ZHAO M, et al Adaptive underwater image enhancement based on color compensation and fusion[J]. Signal, Image and Video Processing, 2023, 17 (5): 2201- 2210 doi: 10.1007/s11760-022-02435-5
16	HUANG Y, YUAN F, XIAO F, et al Underwater image enhancement based on zero-reference deep network[J]. IEEE Journal of Oceanic Engineering, 2023, 48 (3): 903- 924 doi: 10.1109/JOE.2023.3245686
17	PARK C W, EOM I K Underwater image enhancement using adaptive standardization and normalization networks[J]. Engineering Applications of Artificial Intelligence, 2024, 127: 107445 doi: 10.1016/j.engappai.2023.107445
18	GUAN F, LU S, LAI H, et al AUIE-GAN: adaptive underwater image enhancement based on generative adversarial networks[J]. Journal of Marine Science and Engineering, 2023, 11 (7): 1476 doi: 10.3390/jmse11071476
19	梁秀满, 赵治刚, 于海峰, 等基于颜色线模型和多尺度融合的水下图像增强[J]. 控制与决策, 2025, 40 (1): 162- 169 LIANG Xiuman, ZHAO Zhigang, YU Haifeng, et al Underwater image enhancement based on color-line model and multi-scale fusion[J]. Control and Decision, 2025, 40 (1): 162- 169
20	温佩芝, 陈君谋, 肖雁南, 等基于生成式对抗网络和多级小波包卷积网络的水下图像增强算法[J]. 浙江大学学报: 工学版, 2022, 56 (2): 213- 224 WEN Peizhi, CHEN Junmou, XIAO Yannan, et al Underwater image enhancement algorithm based on GAN and multi-level wavelet CNN[J]. Journal of Zhejiang University: Engineering Science, 2022, 56 (2): 213- 224
21	ZHONG Z, LIN Z Q, BIDART R, et al. Squeeze-and-attention networks for semantic segmentation [C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 13065–13074.
22	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[J]. Advances in Neural Information Processing Systems, 2017, 30, 6000 – 6010.
23	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 doi: 10.1109/TIP.2021.3051462
24	HU J, SHEN L, SUN G. Squeeze-and-excitation networks [C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 7132–7141.
25	ZHU J Y, PARK T, ISOLA P, et al. Unpaired image-to-image translation using cycle-consistent adversarial networks [C]// IEEE International Conference on Computer Vision. Venice: IEEE, 2017: 2242–2251.
26	LI C, GUO J, GUO C Emerging from water: underwater image color correction based on weakly supervised color transfer[J]. IEEE Signal Processing Letters, 2018, 25 (3): 323- 327 doi: 10.1109/LSP.2018.2792050
27	ISLAM M J, XIA Y, SATTAR J Fast underwater image enhancement for improved visual perception[J]. IEEE Robotics and Automation Letters, 2020, 5 (2): 3227- 3234 doi: 10.1109/LRA.2020.2974710
28	LI C, GUO C, REN W, et al An underwater image enhancement benchmark dataset and beyond[J]. IEEE Transactions on Image Processing, 2019, 29: 4376- 4389
29	LI H, LI J, WANG W. A fusion adversarial underwater image enhancement network with a public test dataset [EB/OL]. [2024–03–02]. https://arxiv.org/abs/1906.06819v2.
30	ZUIDERVELD K. Contrast limited adaptive histogram equalization [M]// Graphics gems. Amsterdam: Elsevier, 1994: 474–485.
31	DREWS P JR, DO NASCIMENTO E, MORAES F, et al. Transmission estimation in underwater single images [C]// IEEE International Conference on Computer Vision Workshops. Sydney: IEEE, 2013: 825–830.
32	FU X, ZHUANG P, HUANG Y, et al. A retinex-based enhancing approach for single underwater image [C]// IEEE International Conference on Image Processing. Paris: IEEE, 2014: 4572–4576.
33	PENG Y T, COSMAN P C Underwater image restoration based on image blurriness and light absorption[J]. IEEE Transactions on Image Processing, 2017, 26 (4): 1579- 1594 doi: 10.1109/TIP.2017.2663846
34	LIU S, FAN H, LIN S, et al Adaptive learning attention network for underwater image enhancement[J]. IEEE Robotics and Automation Letters, 2022, 7 (2): 5326- 5333 doi: 10.1109/LRA.2022.3156176

[1]	李灿林,张文娇,邵志文,马利庄,王新玥. 基于Trans-nightSeg的夜间道路场景语义分割方法[J]. 浙江大学学报(工学版), 2024, 58(2): 294-303.
[2]	张剑钊,郭继昌,汪昱东. 基于融合逆透射率图的水下图像增强算法[J]. 浙江大学学报(工学版), 2023, 57(5): 921-929.
[3]	廖嘉文,庞彦伟,聂晶,孙汉卿,曹家乐. 基于特征融合和一致性损失的双目低光照增强[J]. 浙江大学学报(工学版), 2023, 57(12): 2456-2466.
[4]	成宏达,骆海明,夏庆超,杨灿军. 基于改进γ-CLAHE算法的水下机器人图像识别[J]. 浙江大学学报(工学版), 2022, 56(8): 1648-1655.
[5]	温佩芝,陈君谋,肖雁南,温雅媛,黄文明. 基于生成式对抗网络和多级小波包卷积网络的水下图像增强算法[J]. 浙江大学学报(工学版), 2022, 56(2): 213-224.
[6]	程训,余建波. 基于机器视觉的加工刀具磨损监测方法[J]. 浙江大学学报(工学版), 2021, 55(5): 896-904.
[7]	雍子叶,郭继昌,李重仪. 融入注意力机制的弱监督水下图像增强算法[J]. 浙江大学学报(工学版), 2021, 55(3): 555-562.
[8]	王万良,杨小涵,赵燕伟,高楠,吕闯,张兆娟. 采用卷积自编码器网络的图像增强算法[J]. 浙江大学学报(工学版), 2019, 53(9): 1728-1740.
[9]	刘涛, 赵巨峰, 徐之海, 冯华君, 陈慧芳. 基于卡尔曼滤波的红外图像增强算法[J]. J4, 2012, 46(8): 1534-1539.
[10]	张赞超, 夏顺仁. 全自动尿液图像识别技术[J]. J4, 2009, 43(5): 832-838.
[11]	张赞超夏顺仁段会龙. 基于细胞神经网络的尿沉渣图像分割[J]. J4, 2008, 42(12): 2139-2144.
[12]	喻琪夏顺仁丛卫华张峰山杨晓帆. 基于小波系数相关性和模糊理论的声纳图像处理[J]. J4, 2008, 42(12): 2151-2155.
[13]	张懿刘旭李海峰. 适应图像直方图均衡算法[J]. J4, 2007, 41(4): 630-633.

Viewed

Full text

Abstract

Cited

Shared

Discussed