散射光补偿结合色彩保持与对比度均衡的水下图像增强

宁泽萌; 林森; 李兴然

doi:10.11993/j.issn.2096-3920.2023-0131

散射光补偿结合色彩保持与对比度均衡的水下图像增强

doi: 10.11993/j.issn.2096-3920.2023-0131

宁泽萌^1,,
林森^1, ,,
李兴然^2,

1.
沈阳理工大学自动化与电气工程学院, 辽宁沈阳, 110159
2.
河海大学河海里尔学院, 江苏南京, 210098

基金项目: 国家重点研发计划项目(2018YFB1403303), 辽宁省教育厅高等学校基本科研项目(LJKMZ20220615).

详细信息

通讯作者:
林　森(1980-), 男, 博士, 副教授, 主要研究方向为机器视觉、深度学习、模式识别

中图分类号: TJ630
计量
- 文章访问数: 10
- HTML全文浏览量: 3
- PDF下载量: 1
- 被引次数: 0
出版历程
- 收稿日期: 2023-10-22
- 修回日期: 2023-12-17
- 录用日期: 2024-01-05
- 网络出版日期: 2024-02-07

Scattered Light Compensation Combined with Color Preservation and Contrast Balance for Underwater Image Enhancement

NING Zemeng^1
,,
LIN Sen^{1
, ,},
LI Xingran^2
,

1.
School of Automation and Electrical Engineering, Shenyang Ligong University, Shenyang 110159, China
2.
Hohai-Lille College, Hohai University, Nanjing 210098, China

摘要

摘要: 针对水下图像存在颜色偏差、对比度低及模糊等问题, 提出散射光补偿结合色彩保持与对比度均衡的水下图像增强方法。首先, 利用相对总变分模型分离图像结构层和纹理层。其中, 设计基于RGB空间映射的补偿系数矩阵校正结构层色偏, 并通过滤波分离与融合增强纹理层, 以防止图像的初始特征丢失, 增强后的纹理层与结构层叠加得到第一层输出; 然后, 在对比度均衡模块中, 基于空间变换进行保色对比度限制自适应直方图均衡化, 进一步提高对比度和亮度。最后, 将双层增强结构的结果图融合得到输出图像。通过在不同数据集上与各种经典及新颖的算法对比, 验证本文方法在平衡色差, 增强细节及去模糊等方面均具有更好效果, 对于水下无人系统视觉任务具有实际应用价值。
- 水下图像增强 /
- 色偏校正 /
- 纹理增强 /
- 保色直方图均衡化
Abstract: Aiming at the problems of color deviation, low contrast and blurring in underwater images, an underwater image enhancement method based on scattering light compensation combined with color preservation and contrast balance is proposed. Firstly, the relative total variational model is used to separate the structure and texture layer. Among them, the color deviation of the structural layer is corrected by defining a compensation coefficient error matrix based on the RGB spatial mapping relationship, and the texture layer is enhanced by filtering separation and fusion to prevent the initial feature loss of the image. Besides, color preservation-contrast limiting adaptive histogram equalization based on the spatial transformation is performed to further improve the contrast and brightness. Finally, the double-stream enhanced results are fused to obtain the output. It is verified by various evaluations on different datasets that the proposed method has better performance in balancing color deviation, enhancing details, and dehazing, which has practical application value in underwater computer vision tasks.
- Underwater image enhancement /
- Color correction /
- Texture enhancement /
- Color preservation-histogram equalization

HTML全文

图 1 所提方法流程图

Figure 1. Flow chart of the proposed method

下载: 全尺寸图片幻灯片

图 2 相对总变差模型分解图像

Figure 2. Decomposition results of RTV

下载: 全尺寸图片幻灯片

图 3 颜色校正算法结果

Figure 3. Color correction results

下载: 全尺寸图片幻灯片

图 4 CP-CLAHE效果展示

Figure 4. Enhancement and histogram results of the CP-CLAHE

下载: 全尺寸图片幻灯片

图 5 定性评价

Figure 5. Qualitative evaluation

下载: 全尺寸图片幻灯片

图 6 特征点匹配示例

Figure 6. Example of feature point matching

下载: 全尺寸图片幻灯片

图 7 视频单帧图像增强结果

Figure 7. Enhancement results of single frames in video

下载: 全尺寸图片幻灯片

表 1 CIEDE2000的评价结果

Table 1. Evaluation results for CIEDE2000

方法	D10	Z33	T6000	T8000	TS1	W60	W80	平均值
raw	22.189	23.262	23.277	29.081	24.341	22.07	26.172	24.342
UDCP	24.407	25.075	22.401	30.886	24.644	21.167	27.934	25.216
Fusion	19.631	18.947	16.458	19.517	19.342	20.186	24.544	19.804
GDCP	25.489	24.125	22.78	22.906	21.141	23.036	30.847	24.332
Hybrid	22.421	23.031	20.938	22.994	20.508	20.815	22.785	21.927
MLLE	20.604	26.136	27.543	19.048	18	22.729	21.578	22.234
CFA&MR	21.217	29.297	21.544	23.643	17.034	20.982	23.915	22.519
UIESS	19.405	21.309	17.692	23.545	21.461	17.672	19.633	20.102
Shallow	20.251	18.857	17.152	27.448	27.622	19.783	25.059	22.31
文中方法	19.339	17.913	18.235	18.814	15.663	19.066	20.815	18.549

下载: 导出CSV

表 2 UIEB和EUVP数据集的评价结果

Table 2. Evaluation results on UIEB and EUVP databases

方法	UIEB				EUVP
方法	SSIM	PSNR	FDUM	NIQE	SSIM	PSNR	FDUM	NIQE
UDCP	0.682	17.065	0.755	2.887	0.631	15.630	0.578	3.637
GDCP	0.599	13.525	0.854	2.921	0.597	14.306	0.682	3.470
Fusion	0.768	20.760	0.669	2.781	0.708	19.419	0.495	3.516
Hybrid	0.641	17.309	1.051	3.245	0.583	16.636	0.790	2.845
MLLE	0.692	16.111	0.948	3.523	0.631	15.462	0.702	2.940
CFA&MR	0.574	18.479	0.495	2.920	0.466	17.613	0.340	3.846
UIESS	0.624	17.811	0.585	4.034	0.687	19.898	0.537	3.985
Shallow	0.534	16.625	0.455	3.726	0.760	20.238	0.459	4.234
文中方法	0.789	20.406	0.951	2.832	0.768	20.255	0.820	3.590

下载: 导出CSV

表 3 平均运行时间

Table 3. Average running time

方法	UDCP	Fusion	GDCP	Hybrid	MLLE	CFA&MR	UIESS	Shallow	文中方法
时间	0.618	2.602	0.382	10.522	0.082	1.581	0.650	0.110	0.365

下载: 导出CSV

表 4 特征点匹配结果

Table 4. Feature point matching results

方法	UDCP	Fusion	GDCP	Hybrid	MLLE	CFA&MR	UIESS	Shallow	文中方法
数量	21	29	37	39	23	41	37	37	44

下载: 导出CSV

参考文献(27)

[1]	王丹, 张子玉, 赵金宝, 等. 基于常散射假设和同态滤波的水下图像增强算法[J]. 水下无人系统学报, 2021, 29(2): 210-217. Wang Dan, Zhang Ziyu, Zhao Jinbao, et al. Underwater image enhancement algorithm based on constant scattering assumption and homomorphic filtering[J]. Journal of Unmanned Undersea Systems, 2021, 29(2): 210-217.
[2]	姚鹏, 刘玉会. 基于UNDERWATER-CUT模型的水下图像增强算法[J]. 水下无人系统学报, 2022, 30(5): 605-611. Yao Peng, Liu Yuhui. Underwater image enhancement based on UNDERWATER-CUT model[J]. Journal of Unmanned Undersea Systems, 2022, 30(5): 605-611.
[3]	Lei X, Wang H, Shen J, et al. Underwater image enhancement based on color correction and complementary dual image multi-scale fusion[J]. Applied Optics, 2022, 61: 5304-14. doi: 10.1364/AO.456368
[4]	He K, Sun J, Tang X. Single image haze removal using dark channel prior[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33: 2341-52. doi: 10.1109/TPAMI.2010.168
[5]	Hou G, Li J, Wang G, et al. A novel dark channel prior guided variational framework for underwater image restoration[J]. Journal of Visual Communication and Image Representation, 2020, 66: 102732. doi: 10.1016/j.jvcir.2019.102732
[6]	Liang Z, Ding X, Wang Y, et al. GUDCP: Generalization of underwater dark channel prior for underwater image restoration[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2022, 32(7): 4879-84. doi: 10.1109/TCSVT.2021.3114230
[7]	Zhou J, Yan Y, Chu W, et al. Underwater image restoration via backscatter pixel prior and color compensation[J]. Engineering Applications of Artificial Intelligence, 2022, 111: 104785. doi: 10.1016/j.engappai.2022.104785
[8]	Zhang W, Zhuang P, Sun 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
[9]	Li X, Hou G, Li K, et al. Enhancing underwater image via adaptive color and contrast enhancement, and denoising[J]. Engineering Applications of Artificial Intelligence, 2022, 111: 104759. doi: 10.1016/j.engappai.2022.104759
[10]	Zhang W, Pan X, Xi X, et al. Color Correction and adaptive contrast enhancement for underwater image enhancement[J]. Computers & Electrical Engineering, 2021, 91: 106981.
[11]	Zhuang P, Li C, Wu J, et al. Bayesian retinex underwater image enhancement[J]. Engineering Applications of Artificial Intelligence, 2021, 101: 104171. doi: 10.1016/j.engappai.2021.104171
[12]	Zhuang P, Ding X. Underwater image enhancement using an edge-preserving filtering retinex algorithm[J]. Multimed Tools Appl, 2020, 79: 17257-77. doi: 10.1007/s11042-019-08404-4
[13]	Zhuang P, Wu J, Porikli F, et al. Underwater image enhancement with hyper-laplacian reflectance priors[J]. IEEE Transactions on Image Processing, 2022, 31: 5442-55. doi: 10.1109/TIP.2022.3196546
[14]	Wang Y, Zhang J, Cao Y, et al. A deep CNN method for underwater image enhancement[C]//24th IEEE International Conference on Image Processing(ICIP). Beijing, China: IEEE, 2017: 1382-86.
[15]	Li C, Guo C, Ren W, et al. An underwater image enhancement benchmark dataset and beyond[J]. IEEE Transactions on Image Processing, 2020, 29: 4376-89. doi: 10.1109/TIP.2019.2955241
[16]	Yang M, Hu K, Du Y, et al. Underwater image enhancement based on conditional generative adversarial network[J]. Signal Processing:Image Communication, 2020, 81: 115723. doi: 10.1016/j.image.2019.115723
[17]	林森, 迟凯晨, 唐延东. 基于复原结构与增强纹理融合的水下图像清晰化[J]. 控制与决策, 2022, 37(3): 635-644. Lin Sen, Chi Kaichen, Tang Yandong. Underwater image sharpening based on fusion of restored structure and enhanced texture[J]. Journal of Control and Decision, 2022, 37(3): 635-644.
[18]	Zhou J, We X, Shi J, et al. Underwater image enhancement method with light scattering characteristics[J]. Computers & Electrical Engineering, 2022, 100: 107898.
[19]	Weng C C, Chen H, Fuh C S. A Novel automatic white balance method for digital still cameras[J]. IEEE International Symposium on Circuits and Systems, 2005, 4: 3801-04.
[20]	Lin S, Zhang R, Ning Z, et al. TCRN: A two-step underwater image enhancement network based on triple-color space feature reconstruction[J]. Journal of Marine Science and Engineering, 2023, 11(6): 1221. doi: 10.3390/jmse11061221
[21]	Peng Y, Cao K, Cosman PC. Generalization of the dark channel prior for single image restoration[J]. IEEE Transactions on Image Processing, 2018, 27: 2856-68. doi: 10.1109/TIP.2018.2813092
[22]	Ancuti C O, Ancuti C, De Vleeschouwer, et al. Color balance and fusion for underwater image enhancement[J]. IEEE Transactions on Image Processing, 2018, 27: 379-393. doi: 10.1109/TIP.2017.2759252
[23]	Li C, Tang S, Kwan H, et al. Color correction based on CFA and enhancement based on retinex with dense pixels for underwater images[J]. IEEE Access, 2020, 8: 155732-41. doi: 10.1109/ACCESS.2020.3019354
[24]	Li X, Hou G, Tan L, et al. A hybrid framework for underwater image enhancement[J]. IEEE Access, 2020, 8: 197448-62. doi: 10.1109/ACCESS.2020.3034275
[25]	Naik A, Swarnakar A, Mittal K. Shallow-uwnet: Compressed model for underwater image enhancement[C]// Proceedings of the AAAI Conference on Artificial Intelligence, 2021: 15853-54.
[26]	Chen Y, Pei S. Domain adaptation for underwater image enhancement via content and style separation[J]. IEEE Access, 2022, 10: 90523-34. doi: 10.1109/ACCESS.2022.3201555
[27]	Yang N, Zhong Q, Li K, et al. A reference-free underwater image quality assessment metric in frequency domain[J]. Signal Process, 2021, 94: 116218.