基于多尺度编码-解码神经网络的图像去雾算法

doi:10.13229/j.cnki.jdxbgxb.20231030

摘要/Abstract

摘要：

针对有雾的场景中图像采集系统采集到的图像会出现细节缺失、色彩暗淡、亮度降低的问题，在一体化去雾网络（AOD）理论的基础上提出一种多尺度编码-解码神经网络（MSAOD）模型进行图像去雾。本文网络模型分为3个模块：①预处理模块，将输入图像分为2部分进行预处理；②主干模块，通过多尺度编码器-解码器对第1部分的输出进行特征提取；③后处理模块，对特征图进行映射操作。通过训练得到去雾图像，实验结果表明，本文方法要优于主流的深度学习和传统方法的图像去雾效果，去雾后的图像在细节、色彩和亮度等方面都有所优化。

关键词: 数字图像处理, 图像去雾, 大气散射模型, 深度学习, 多尺度网络

Abstract:

Aiming at the problems of missing details， dim color and reduced brightness in the images collected by the image acquisition system in the foggy scene， a multi-scale encoding decoding neural network （MSAOD） model was proposed based on the AOD theory for image dehazing. The proposed network model was divided into three modules. The first module is the preprocessing module， which divides the input image into two parts for preprocessing. The second module is the backbone module， which extracts the features of the output of the first part through the multi-scale encoder decoder. The third module is the post-processing module， which maps the feature map. The experimental results show that this method is superior to the mainstream deep learning and traditional methods in image dehazing effect， and the image after dehazing is optimized in detail， color， brightness and so on.

Key words: digital image processing, image dehazing, atmospheric scattering model, deep learning, multi-scale network

中图分类号:

TP399

王勇,边宇霄,李新潮,徐椿明,彭刚,王继奎. 基于多尺度编码-解码神经网络的图像去雾算法[J]. 吉林大学学报(工学版), 2024, 54(12): 3626-3636.

Yong WANG,Yu-xiao BIAN,Xin-chao LI,Chun-ming XU,Gang PENG,Ji-kui WANG. Image dehazing algorithm based on multiscale encoding decoding neural network[J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(12): 3626-3636.

图/表 19

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

图11

图12

图13

图14

表1

图15

表2

图16

表4

参考文献 32

1	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, 2010, 33(12): 2341-2353.
2	Tarel J P, Hautiere N. Fast visibility restoration from a single color or gray level image[C]∥2009 IEEE 12th International Conference on Computer Vision. Kyoto, Japan,2009: 2201-2208.
3	Sulami M, Glatzer I, Fattal R, et al. Automatic recovery of the atmospheric light in hazy images[C]∥2014 IEEE International Conference on Computational Photography (ICCP).Santa Clara, USA,2014: 1-11.
4	Cai B L, Xu X M, Jia K. An end-to-end system for single image haze removal[J].IEEE Transactions on Image Processing, 2016, 25(11): 5187-5198.
5	Qin X, Wang Z L, Bai Y C, et al. FFA-Net: Feature fusion attention network for single image dehazing[C]∥Proceedings of the AAAI Conference on Artificial Intelligence,New York,USA,2020: 11908-11915.
6	顼聪,陶永鹏,朱毅,等.基于图像块高位平面无损压缩的可逆信息隐藏方法[J].东北师大学报: 自然科学版, 2023, 55(3): 101-108.
	Xu Cong, Tao Yong-peng, Zhu Yi. Reversible data hiding method based on image block high bit planes lossless compression[J]. Journal of Northeast Normal University(Natural Science Edition), 2023, 55(3): 101-108.
7	Liu X H, Ma Y R, Shi Z H, et al. Griddehazenet: Attention-based multi-scale network for image dehazing[C]∥Proceedings of the IEEE/CVF International Conference on Computer Vision,Seoul,Korean, 2019: 7314-7323.
8	刘雪,赵辰一,乔双,等. 基于SVD和残差图像的噪声强度估计方法及其在中子图像去噪中的应用[J].东北师大学报: 自然科学版, 2024, 56(3): 62-69.
	Liu Xue, Zhao Chen-yi, Qiao Shuang,et al. Noise intensity estimation method based on SVD and residual images for neutron image denoising[J]. Journal of Northeast Normal University(Natural Science Edition), 2024, 56(3): 62-69.
9	周军,孔菁菁.多策略改进的麻雀搜索算法及其应用[J].东北师大学报: 自然科学版, 2024, 56(4): 81-88.
	Zhou Jun, Kong Jing-jing. Multi-strategy improved sparrow search algorithm and its application[J]. Journal of Northeast Normal University(Natural Science Edition), 2024, 56(4): 81-88.
10	Shit S, Das D K, Ray D N, et al. An encoder‐decoder based CNN architecture using end to end dehaze and detection network for proper image visualization and detection[J]. Computer Animation and Virtual Worlds, 2023, 34(6): No.e2147.
11	Li B, Zhao J J, Fu H. DLT-Net: deep learning transmittance network for single image haze removal[J]. Signal, Image and Video Processing, 2020, 14(6): 1245-1253.
12	Zhang S D, He F Z, Ren W Q, et al. Joint learning of image detail and transmission map for single image dehazing[J]. The Visual Computer, 2020, 36(2): 305-316.
13	Wang A N, Wang W H, Liu J L, et al. AIPNet: Image-to-image single image dehazing with atmospheric illumination prior[J]. IEEE Transactions on Image Processing, 2018, 28(1): 381-393.
14	Roy S, Chaudhuri S S. SIVDSR-Dhaze: Single Image Dehazing with Very Deep Super Resolution Framework and Its Analysis[J].Scientific Visualization, 2022, 14(5): 16-44.
15	Wang C, Shen H Z, Fan F, et al. EAA-Net: a novel edge assisted attention network for single image dehazing[J]. Knowledge-Based Systems, 2021, 228: 107279.
16	Baskar H, Chakravarthy A S, Garg P, et al. NDENet: end-to-end nighttime dehazing and enhancement[J]. International Journal of Computer and Information Engineering, 2022, 16(11): 540-548.
17	Thomas R, Thampi L, Kamal S, et al. Dehazing underwater images using encoder decoder based generic model-agnostic convolutional neural network[C]∥2021 International Symposium on Ocean Technology (SYMPOL), Piscataulay,NJ, 2021: 1-4.
18	Zhao L Q, Zhang Y P, Cui Y. An attention encoder-decoder network based on generative adversarial network for remote sensing image dehazing[J]. IEEE Sensors Journal, 2022, 22(11): 10890-10900.
19	Kim G, Ha S, Kwon J. Adaptive patch based convolutional neural network for robust dehazing[C]∥2018 25th IEEE International Conference on Image Processing (ICIP), Athens, Greece, 2018: 2845-2849.
20	Li C Y, Guo J C, Porikli F, et al. A cascaded convolutional neural network for single image dehazing[J]. IEEE Access, 2018, 6: 24877-24887.
21	Song Y F, Li J, Wang X G, et al. Single image dehazing using ranking convolutional neural network[J]. IEEE Transactions on Multimedia, 2017, 20(6): 1548-1560.
22	You Y, Lu C W, Wang W M, et al. Relative CNN-RNN: learning relative atmospheric visibility from images[J]. IEEE Transactions on Image Processing, 2018, 28(1): 45-55.
23	Chen D D, He M M, Fan Q N, et al. Gated context aggregation network for image dehazing and deraining[C]∥2019 IEEE Winter Conference on Applications of Computer Vision (WACV),Waikoloa Village,USA, 2019: 1375-1383.
24	Liu K, He L Y, Ma S P, et al. A sensor image dehazing algorithm based on feature learning[J]. Sensors, 2018, 18(8): 2606.
25	Xiao J S, Zhou J L, Lei J F, et al. Image hazing algorithm based on generative adversarial networks[J]. IEEE Access, 2019, 8: 15883-15894.
26	Li R D, Pan J S, Li Z H, et al. Single image dehazing via conditional generative adversarial network[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,Salt Lake City,USA, 2018: 8202-8211
27	Liu W, Hou X X, Duan J, et al. End-to-end single image fog removal using enhanced cycle consistent adversarial networks[J]. IEEE Transactions on Image Processing, 2020, 29: 7819-7833.
28	Zhang X H.Research on Remote sensing image de-haze based on GAN[J].Journal of Signal Processing Systems, 2022, 94(3): 305-313.
29	Zheng Y T, Su J, Zhang S, et al. Dehaze-AGGAN: unpaired remote sensing image dehazing using enhanced attention-guide generative adversarial networks[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 1-13.
30	Wang C H, Meng Z J, Xie R L, et al. A single image dehazing algorithm based on cycle-GAN[C]∥Proceedings of the 2019 International Conference on Robotics, Intelligent Control and Artificial Intelligence,Shanghai, China, 2019: 247-251.
31	郭璠. 图像去雾方法和评价及其应用研究[D]. 长沙:中南大学信息科学与工程学院, 2012.
	Guo Pan.Research on image defogging,effect assessment and application[D]. Changsha: School of Information Science and Engineering, Central South University,2012.
32	Narasimhan S G, Nayar S K. Removing weather effects from monochrome images[C]∥Proceedings of the 2001 IEEE Computer society Conference on Computer Vision and Pattern Recognition,Kauai, USA, 2001, 2: No.II.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

方法	PSNR	SSIM
FVR	12.879 6	0.587 2
DCP	13.230 0	0.703 3
AL	18.819 5	0.862 7
ATM	7.977 3	0.694 0
Dehazednet	9.770 0	0.484 6
AOD-Net	19.704 8	0.751 8
本文方法	21.817 0	0.895 0

评分指标	FVR	ATM	DCP	AL	Dehazednet	AOD	MSAOD
Brenner	0.822 8	0.310 1	0.582 5	0.863 1	0.613 2	0.529 1	0.874 1
Laplacian	0.074 9	0.016 3	0.067 2	0.071 6	0.051 8	0.057 8	0.075 9
SMD	0.035 3	0.020 4	0.032 6	0.047 3	0.034 3	0.027 9	0.049 6
SMD2	2.145 8	0.795 7	1.868 3	2.436 1	1.252 3	1.305 4	2.616 5
Variance	4.841 0	2.972 9	7.416 7	8.239 1	5.364 3	6.758 3	8.207 3
Energy	1.059 6	0.307 9	1.049 8	1.053 8	0.742 7	0.509 4	0.793 0
Vollath	4.798 0	2.694 2	6.179 5	8.524 7	5.732 0	6.651 0	8.537 6
Entropy	4.856 3	4.545 4	4.453 8	5.077 8	4.135 8	4.969 7	5.023 1

评分指标	1	2	3	2+3	1+2+3	本文
PSNR	20.286	6.096	20.879	20.286	19.908	22.585
Brenner	0.655	0.678	0.621	0.632	0.938	1.182
Laplacian	0.039	0.041	0.021	0.036	0.052	0.054
Energy	0.821	0.612	0.871	0.686	0.836	0.896

[1]	张磊,焦晶,李勃昕,周延杰. 融合机器学习和深度学习的大容量半结构化数据抽取算法[J]. 吉林大学学报(工学版), 2024, 54(9): 2631-2637.
[2]	李路,宋均琦,朱明,谭鹤群,周玉凡,孙超奇,周铖钰. 基于RGHS图像增强和改进YOLOv5网络的黄颡鱼目标提取[J]. 吉林大学学报(工学版), 2024, 54(9): 2638-2645.
[3]	乔百友,武彤,杨璐,蒋有文. 一种基于BiGRU和胶囊网络的文本情感分析方法[J]. 吉林大学学报(工学版), 2024, 54(7): 2026-2037.
[4]	郭昕刚,何颖晨,程超. 抗噪声的分步式图像超分辨率重构算法[J]. 吉林大学学报(工学版), 2024, 54(7): 2063-2071.
[5]	张丽平,刘斌毓,李松,郝忠孝. 基于稀疏多头自注意力的轨迹kNN查询方法[J]. 吉林大学学报(工学版), 2024, 54(6): 1756-1766.
[6]	孙铭会,薛浩,金玉波,曲卫东,秦贵和. 联合时空注意力的视频显著性预测[J]. 吉林大学学报(工学版), 2024, 54(6): 1767-1776.
[7]	陆玉凯,袁帅科,熊树生,朱绍鹏,张宁. 汽车漆面缺陷高精度检测系统[J]. 吉林大学学报(工学版), 2024, 54(5): 1205-1213.
[8]	李雄飞,宋紫萱,朱芮,张小利. 基于多尺度融合的遥感图像变化检测模型[J]. 吉林大学学报(工学版), 2024, 54(2): 516-523.
[9]	杨国俊,齐亚辉,石秀名. 基于数字图像技术的桥梁裂缝检测综述[J]. 吉林大学学报(工学版), 2024, 54(2): 313-332.
[10]	赵彬,吴成东,张雪娇,孙若怀,姜杨. 基于注意力机制的机械臂目标抓取网络技术[J]. 吉林大学学报(工学版), 2024, 54(12): 3423-3432.
[11]	车娜,朱奕明,赵剑,孙磊,史丽娟,曾现伟. 基于联结主义的视听语音识别方法[J]. 吉林大学学报(工学版), 2024, 54(10): 2984-2993.
[12]	陈岳林,高铸成,蔡晓东. 基于BERT与密集复合网络的长文本语义匹配模型[J]. 吉林大学学报(工学版), 2024, 54(1): 232-239.
[13]	霍光,林大为,刘元宁,朱晓冬,袁梦,盖迪. 基于多尺度特征和注意力机制的轻量级虹膜分割模型[J]. 吉林大学学报(工学版), 2023, 53(9): 2591-2600.
[14]	金小俊,孙艳霞,于佳琳,陈勇. 基于深度学习与图像处理的蔬菜苗期杂草识别方法[J]. 吉林大学学报(工学版), 2023, 53(8): 2421-2429.
[15]	耿庆田,刘植,李清亮,于繁华,李晓宁. 基于一种深度学习模型的土壤湿度预测[J]. 吉林大学学报(工学版), 2023, 53(8): 2430-2436.