基于光照补偿与金字塔融合的水下图像增强方法

岳成海; 徐会希; 吕凤天; 邵刚; 朱宝彤; 尹忠勋

doi:10.11993/j.issn.2096-3920.2024-0082

基于光照补偿与金字塔融合的水下图像增强方法

doi: 10.11993/j.issn.2096-3920.2024-0082

岳成海^{1, 2},
徐会希^{1, 2},
吕凤天^{1, 2},
邵刚^{1, 2},
朱宝彤^{1, 2},
尹忠勋^{1, 2}

1.
中国科学院沈阳自动化研究所机器人学国家重点实验室, 辽宁沈阳, 110016
2.
辽宁省水下机器人重点实验室, 辽宁沈阳, 110169

基金项目: 国家重点研发计划项目(2021YFC2800100); 广东省重点领域研发计划项目(2020B1111010004); 辽宁省自然科学基金(2022-MS-035).

详细信息

作者简介:
岳成海(1989-), 男, 硕士, 助理研究员, 主要研究方向为成像探测与智能信息处理

中图分类号: TP23
计量
- 文章访问数: 9
- HTML全文浏览量: 4
- PDF下载量: 0
- 被引次数: 0
出版历程
- 收稿日期: 2024-05-20
- 修回日期: 2024-07-01
- 录用日期: 2024-07-02
- 网络出版日期: 2025-01-16

Underwater image enhancement method based on light compensation and pyramid fusion

YUE Chenghai^{1, 2},
XU Huixi^{1, 2},
LÜ Fengtian^{1, 2},
SHAO Gang^{1, 2},
ZHU Baotong^{1, 2},
YIN Zhongxun^{1, 2}

1.
State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110016, China
2.
Key Laboratory of Marine Robotics, Liaoning Province, Shenyang 110169, China

摘要

摘要: 水下光学成像存在色偏、散射模糊与亮度不均的问题, 现有的基于深度学习的方法与基于水下成像像质退化模型的图像增强方法仍然存在鲁棒性差的问题。针对上述问题, 文中提出光照补偿与金字塔细节融合的单幅水下图像增强方法, 首先结合全局照度与色彩通道特性在像素级实现光照强度的估计与补偿, 实现各色彩通道的强度校正, 然后以高斯模糊估计图像散射分量并采用多尺度高斯滤波残差法去散射, 最后提出融合边缘增强、自适应Gamma校正及亮度均衡的多图金字塔细节融合亮度均衡方法, 较好地保留图像细节信息的同时, 解决图像亮度不均问题。对比现有方法, 文中方法适应性更好, 在水下图像质量评价指标(UIQM)与水下图像颜色质量评价指标(UCIQE)等方面都具有性能提升的优势。
- 光照补偿 /
- 去散射 /
- 金字塔融合 /
- 亮度均衡 /
- 图像增强
Abstract: The existing enhancement methods based on deep learning and underwater imaging models still have insufficient adaptability due to the problems of color deviation, scattering blur, and uneven brightness in underwater optical imaging. In response to the above issues, this article proposes a single underwater image enhancement method that combines lighting compensation and pyramid detail fusion. Firstly, the global illumination and color channel characteristics are combined to estimate and compensate lighting intensity at pixel level, achieving intensity correction for each color channel. Then, Gaussian blur is used to estimate the scattered components of the image and multi-scale gaussian filter residual method is used to remove scattering. Finally, a multi image pyramid detail fusion brightness equalization method that combines edge enhanced image, adaptive Gamma corrected image, and brightness balanced image is proposed, which effectively preserves image detail information while solving the problem of uneven brightness in the image. Subjective evaluation and objective analysis have demonstrated the effectiveness of the method proposed in this paper.
- Light compensation /
- Descattering /
- Pyramid fusion /
- Brightness balance /
- Image enhancement

HTML全文

图 1 水下光学成像

Figure 1. Underwater optical imaging

下载: 全尺寸图片幻灯片

图 2 算法流程图

Figure 2. Algorithm flow chart

下载: 全尺寸图片幻灯片

图 3 光照补偿测试图

Figure 3. Diagram of light compensation test

下载: 全尺寸图片幻灯片

图 4 白平衡测试

Figure 4. White balance test

下载: 全尺寸图片幻灯片

图 5 去散射测试

Figure 5. Descattering test

下载: 全尺寸图片幻灯片

图 6 光照均衡测试

Figure 6. Light balance test

下载: 全尺寸图片幻灯片

图 7 光照均衡测试

Figure 7. 对比测试Compared images

下载: 全尺寸图片幻灯片

图 8 去雾测试

Figure 8. Defogging test

下载: 全尺寸图片幻灯片

表 1 评价指标对比

Table 1. Comparison of evaluation indicators

方法	UCIQE	UICM	UISM	UIConM	UIQM
RAW	0.371 2	9.909 7	1.290 9	−0.153 0	0.113 4
CLAHE	0.382 1	13.420 0	1.290 3	−0.190 8	0.077 3
RETINEX	0.379 1	16.559 9	1.531 4	−0.188 8	0.244 0
CBF	0.372 4	17.488 3	1.518 9	−0.198 2	0.232 9
CCIA	0.365 6	15.231 0	1.206 0	−0.190 9	0.103 0
Shallow-UWnet	0.341 9	7.152 7	0.514 1	−0.198 6	−0.356 5
SyreaNet	0.365 6	15.852 5	1.244 1	−0.199 0	0.102 6
MLLE	0.369 5	13.038 4	2.249 8	−0.184 9	0.370 9
WWPF	0.370 5	13.876 5	1.965 6	−0.175 6	0.343 8
文中算法	0.3879	16.400 1	0.847 8	−0.092 2	0.383 0

下载: 导出CSV

参考文献(24)

[1]	ZHANG M, PENG J. Underwater image restoration based on a new underwater image formation model[J]. IEEE Access, 2018, 6: 58634-58644. doi: 10.1109/ACCESS.2018.2875344
[2]	LEE H S, MOON S W, EOM I K. Underwater image enhancement using successive color correction and superpixel dark channel prior[J]. Symmetry, 2020, 12(8): 1220. doi: 10.3390/sym12081220
[3]	李玉鑫, 梁天全, 于会山, 等. 融合暗通道与Retinex算法的水下图像复原研究[J]. 计算机仿真, 2024, 41(3): 162-167, 172. doi: 10.3969/j.issn.1006-9348.2024.03.031 LI Y X, LIANG T Q, YU H S, et al. research on underwater image restoration combining dark channel and Retinex algorithm[J]. Computer Simulation, 2024, 41(3): 162-167, 172. doi: 10.3969/j.issn.1006-9348.2024.03.031
[4]	ZHUANG P X, WU J M, PORIKLI F, et al. Underwater image enhancement with hyper-laplacian reflectance priors[J]. IEEE Transactions on Image Processing, 2022, 31: 5442-5455. doi: 10.1109/TIP.2022.3196546
[5]	ZHANG W, ZHUANG P, SUN H H, et al. Underwater image enhancement via minimal color loss and locally adaptive contrast enhancement[J]. IEEE Transactions on Image Processing, 2022, 31: 3997-4010. doi: 10.1109/TIP.2022.3177129
[6]	ZHANG W D, ZHOU L, ZHUANG P X, et al. Underwater image enhancement via weighted wavelet visual perception fusion[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2024, 34(4): 2469-2483. doi: 10.1109/TCSVT.2023.3299314
[7]	张微微, 祝开艳. 基于图像融合的低光照水下图像增强[J]. 计算技术与自动化, 2023, 42(04): 85-92. ZHANG W W, ZHU K Y. Low-light underwater image enhancement based on image fusion [J]. Computing technology and automation, 2023, 42 (04): 85-92.
[8]	唐军, 秦艳霞, 林玲, 等. 多尺度融合与细节突显的水下视觉图像增强算法[J/OL]. 机械科学与技术, 1-8[2024-05-04].https://doi.org/10.13433/j.cnki.1003-8728.20240044. TANG J, QIN Y X, LIN L, et al.Underwater visual image enhancement algorithm based on multi-scale fusion and detail highlighting [J/OL]. Mechanical science and technology, 1-8. [2024-05-04].https://doi.org/10.13433/j.cnki.1003-8728.20240044.
[9]	张薇, 郭继昌. 基于白平衡和相对全变分的低照度水下图像增强[J]. 激光与光电子学进展, 2020, 57(12): 213-220. ZHANG W, GUO J C. Low-illumination underwater image enhancement based on white balance and relative total variation [J]. Progress in laser and optoelectronics, 2020, 57 (12): 213-220.
[10]	王悦, 范慧杰, 刘世本, 等. 基于多尺度注意力和对比学习的水下图像增强[J]. 激光与光电子学进展, 2024, 61(04): 559-567. WANG Y, FAN H J, LIU S B, etal. Underwater image enhancement based on multi-scale attention and contrast learning [J]. Progress in laser and optoelectronics, 2024, 61 (4): 559-567.
[11]	陈鑫, 钱旭, 周佳加, 等. 基于水下场景先验的水下图像增强方法研究[J]. 应用科技, 2024, 51(02): 56-65. CHEN X, QIAN X, ZHOU J J, et al. Research on underwater image enhancement method based on underwater scene prior[J]. Application Technology, 2024, 51 (02): 56-65.
[12]	ZHANG T, LI Y, TAKAHASHI S. Underwater image enhancement using improved generative adversarial network[J]. Concurrency and Computation Practice and Experience, 2020(3): 5841.
[13]	李微. 基于改进U-Net网络的水下图像增强[D]. 哈尔滨工程大学, 2021.
[14]	GUO YC, LI H Y, ZHUANG P X, et al.Underwater image enhancement using a multiscale dense generative adversarial network[J]. IEEE Journal of Oceanic Engineering, 2020, 43(3): 862-870.
[15]	FU X Y, C X Y. Underwater image enhancement with global-local networks and compressed-histogram equalization[J]. Signal Processing-Image Communication, 2020, 86: 115892.
[16]	Naik A, Swarnakar A, Mittal K. Shallow-UWnet: compressed model for underwater image enhancement[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2021, 35(18): 15853-15854. doi: 10.1609/aaai.v35i18.17923
[17]	WEN J J, CUI J Q, ZHAO Z J, et al. Syreanet: A physically guided underwater image enhancement framework integrating synthetic and real images[C]//IEEE International Conference on Robotics and Automation(ICRA), London, England:IEEE, 2023: 5177-5183.
[18]	YING Z Q, LI G, REN Y R, et al. A new image contrast enhancement algorithm using exposure fusion framework[J].Computer Analysis of Images and Patterns, 2017(10425): 36-46.
[19]	FU X Y, ZHUANG P X, HUANG Y, et al. A retinex-based enhancing approach for single underwater image[C]//IEEE International Conference on Image Processing(ICIP), Paris, France:IEEE, 2014: 4572-4576.
[20]	DHIVYA R, PRAKASH R, MOHANRAJ M. Color balance and fusion for underwater image enhancement[J]. Digital Image Processing, 2019, 11: 25-29.
[21]	ZHANG W H, LI G, YING Z Q. A new underwater image enhancing method via color correction and illumination adjustment[C]//IEEE Visual Communications and Image Processing(VCIP), St. Petersburg, USA:IEEE, 2017: 1-4.
[22]	LI C Y, GUO C L, REN W Q, et al. An underwater image enhancement benchmark dataset and beyond[J].IEEE Transactions on Image Processing, 2020, 29: 4376-4389.
[23]	PANETTA K, GAO C, AGAIAN S. Human-visual-system-inspired underwater image quality measures[J].IEEE Journal of Oceanic Engineering, 2016, 41(3): 541-551.
[24]	YANG M, SOWMYA A. An underwater color image quality evaluation metric[J].IEEE Transactions on Image Processing, 2015, 24(12): 6062-6071.