基于循环生成对抗网络的图像去雾算法

doi:10.19678/j.issn.1000-3428.0059926

摘要/Abstract

摘要： 大气散射模型与有雾图像及对应清晰图像间的映射模型不适配，导致使用大气散射模型进行图像去雾处理时，图像存在颜色偏差、纹理细节粗糙等问题。基于模拟生物视觉系统的反馈原理，提出一种端到端的循环生成对抗网络算法，以解决误差累积造成的去雾图像低质的问题。通过生成模块将循环神经网络的隐藏状态作为反馈信息，以指导低级模糊特征信息生成更加丰富的高级特征。循环结构能够保证先前的网络层可以使用到后面网络层的高级特征信息，从而减少误差累积。此外，该算法能够根据判别模块的损失来评估重建图像的质量。实验结果表明，与GCANet算法相比，所提算法在SOTS测试集上的平均峰值信噪比和结构相似性，在室内分别提升3.41%和0.57%，在室外分别提升3.48%和1.39%，且在真实世界的数据集上进行图像去雾后，在视觉上避免了颜色失真和光晕问题。

关键词: 图像去雾, 循环卷积神经网络, 生成对抗网络, 编码-解码模式, 反馈连接

Abstract: In image dehazing tasks, sometimes the atmospheric scattering model does not match the model that maps hazy images to clear images, as a result, there are some problems in image defogging using atmospheric scattering model, such as color deviation, rough texture details and so on.To address the accumulated errors and improve the quality of dehazed images, this paper proposes an image dehazing algorithm using an end-to-end recurrent generative adversarial network by simulating the feedback mechanism of biological visual systems.The hidden state of the recurrent neural network in the generative module acts as feedback information to guide the low-level hazy feature information to generate richer high-level features.The recurrent structure ensures that the previous network layer can access the high-level feature information of the next network layer, thus reducing error accumulation.Finally, the quality of reconstructed image is judged according to the loss of the discriminative module.Experimental results on the SOTS test set show that compared with the GCANet-based algorithm, the proposed algorithm provides a 3.41% increase in average Peak Signal-to-Noise Ratio (PSNR) and a 0.57% increase in Structural Similarity (SSIM) in the indoor environment.For the outdoor environment, the proposed algorithm provides a 3.48% increase in PSNR and a 1.39% increase in SSIM.After image defogging on real-world data sets, the problems of color distortion and halo are avoided visually.

Key words: image dehazing, Recurrent Neural Network(RNN), Generative Adversarial Network (GAN), encoding-decoding mode, feedback connection

中图分类号:

TP391.41

郭梦琰, 张娟, 刘巧红, 蔡立志. 基于循环生成对抗网络的图像去雾算法[J]. 计算机工程, 2022, 48(3): 280-287.

GUO Mengyan, ZHANG Juan, LIU Qiaohong, CAI Lizhi. Image Dehazing Algorithm Based on Recurrent Generative Adversarial Network[J]. Computer Engineering, 2022, 48(3): 280-287.

https://www.ecice06.com/CN/Y2022/V48/I3/280

图/表 11

20220401123751

20220401123754

20220401123758

20220401123802

20220401123805

20220401123808

20220401123812

20220401123815

20220401123820

20220401123823

20220401123827

参考文献

[1] 刘尚旺, 崔智勇, 李道义.基于Unet网络多任务学习的遥感图像建筑地物语义分割[J].国土资源遥感, 2020, 32(4):74-83. YANG L S W, CUI Z Y, LI D Y.Semantic segmentation of building features in remote sensing images based on unet multi-task learning[J].Remote Sensing for Land and Resources, 2020, 32(4):74-83.(in Chinese)
[2] 董鹏曙, 向龙, 谢幼才, 等.基于运动补偿的动目标检测处理方法[J].探测与控制学报, 2020, 42(6):29-34. DONG P S, XIANG L, XIE Y C, et al.MTD process metrology based on motion compensation[J].Journal of Detection and Control, 2020, 42(6):29-34.(in Chinese)
[3] GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al.Generative adversarial nets[C]//Proceedings of the 27th International Conference on Neural Information Processing Systems.New York, USA:ACM Press, 2014:2672-2680.
[4] JU M Y, DING C, ZHANG D Y, et al.Gamma-correction-based visibility restoration for single hazy images[J].IEEE Signal Processing Letters, 2018, 25(7):1084-1088.
[5] 董静薇, 赵春丽, 海博.融合同态滤波和小波变换的图像去雾算法研究[J].哈尔滨理工大学学报, 2019, 24(1):66-70. DONG J W, ZHAO C L, HAI B.Research on image De-fog algorithm based on fusion homomorphic filtering and wavelet transform[J].Journal of Harbin University of Science and Technology, 2019, 24(1):66-70.(in Chinese)
[6] LI M D, LIU J Y, YANG W H, et al.Structure-revealing low-light image enhancement via robust retinex model[J].IEEE Transactions on Image Processing, 2018, 27(6):2828-2841.
[7] WANG W C, YUAN X H, WU X J, et al.Fast image dehazing method based on linear transformation[J].IEEE Transactions on Multimedia, 2017, 19(6):1142-1155.
[8] 刘策, 杨燕.基于自适应小波融合的单幅图像去雾算法[J].光电子·激光, 2020, 31(3):318-325. LIU C, YANG Y.Single image dehazing algorithm based on adaptive wavelet fusion[J].Journal of Optoelectronics·Laser, 2020, 31(3):318-325(in Chinese)
[9] MCCARTNEY E J.Optics of the atmosphere:scattering by molecules and particles[J].John Wiley and Sons, 1976, 10(2):461-470.
[10] HE K M, SUN J, TANG X O, et al.Single image haze removal using dark channel prior[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 33(12):2341-2353.
[11] BERMAN D, TREIBITZ T, AVIDAN S.Non-local image dehazing[C]//Proceedings of IEEE Conference on Computer Vsion and Pattern Recognition.Washington D.C., USA:IEEE Press, 2016:1674-1682.
[12] 于坤, 焦青亮, 刘子龙, 等.基于中智学的非局部先验图像去雾算法[J].光学技术, 2020, 46(4):476-482. YU K, JIAO Q L, LIU Z L, et al.Non-local prior image dehazing algorithm based on neutrosophy[J].Optical Technique, 2020, 46(4):476-482.(in Chinese)
[13] 王殿伟, 闫伟超, 刘颖, 等.基于暗原色先验的快速单幅图像去雾算法[J].计算机应用研究, 2018, 35(12):3836-3840. WANG D W, YAN W C, LIU Y, et al.Fast single image dehazing algorithm based on dark channel prior[J].Application Research of Computers, 2018, 35(12):3836-3840.(in Chinese)
[14] CAI B L, XU X M, JIA K, et al.Dehazenet:an end-to-end system for single image haze removal[J].IEEE Transactions on Image Processing, 2016, 25(11):5187-5198.
[15] ZHANG H, PATEL V M, et al.Densely connected pyramid dehazing network[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2018:3194-3203.
[16] CHEN D D, HE M M, FAN Q N, et al.Gated context aggregation network for image dehazing and deraining[C]//Proceedings of 2019 IEEE Winter Conference on Applications of Computer Vision.Washington D.C., USA:IEEE Press, 2019:1375-1383.
[17] LIU X H, MA Y R, SHI Z H, et al.Griddehazenet:attention-based multi-scale network for image dehazing[C]//Proceedings of IEEE International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2019:7314-7323.
[18] DENG Z J, Z L, HU X W, et al.Deep multi-model fusion for single-image dehazing[C]//Proceedings of IEEE International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2019:2453-2462.
[19] 任敏敏.图像融合的循环神经网络去雾算法[J].小型微型计算机系统, 2020, 41(7):1513-1518. REN M M.Recurrent network for Image Dehazing based on image fusion scheme[J].Journal of Chinese Mini-Micro Computer Systems, 2020, 41(7):1513-1518.(in Chinese)
[20] 刘波宁, 翟东海.基于双鉴别网络的生成对抗网络图像修复方法[J].计算机应用, 2018, 38(12):3557-3562, 3595. LIU B N, ZHAI D H.Image completion method of generative adversarial networks based on two discrimination networks[J].Journal of Computer Applications, 2018, 38(12):3557-3562, 3595.(in Chinese)
[21] 斯蒋芸, 谭宁, 张海, 等.基于条件生成对抗网络的咬翼片图像分割[J].计算机工程, 2019, 45(4):223-227. SI J Y, TAN N, ZHANG H, et al.Bitewing radiography image segmentation based on conditio-nal generative adversarial network[J].Computer Engineering, 2019, 45(4):223-227.(in Chinese)
[22] ARJOVSKY M, CHINTALA S, BOTTOU L.Wasserstein gan[EB/OL].[2020-10-01] https://arxiv.org/abs/1701.07875v3.
[23] LI B Y, REN W Q, FU D P, et al.Benchmarking single-image dehazing and beyond[J].IEEE Transactions on Image Processing, 2018, 28(1):492-505.
[24] FATTAL R.Dehazing using color-lines[J].ACM Transactions on Graphics, 2014, 34(1):1-14.
[25] SILBERMAN N, HOIEM D, KOHLI P, et al.Indoor segmentation and support inference from RGBD images[C]//Proceedings of European Conference on Computer Vision.Berlin, Germany:Springer, 2012:746-760.
[26] SCHARSTEIN D, SZELISKI R.High-accuracy stereo depth maps using structured light[C]//Proceedings of 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Washington D.C., USA:IEEE Press, 2003:11-17.
[27] LI B Y, PENG X L, WANG Z Y, et al.Aod-net:all-in-one dehazing network[C]//Proceedings of IEEE International Conference on Computer Vision, Washington D.C., USA:IEEE Press, 2017:4770-4778.
[28] REN W Q, MA L, ZHANG J W, et al.Gated fusion network for single image dehazing[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2018:3253-3261.
[29] HE K M, ZHANG X Y, REN S Q, et al.Delving deep into rectifiers:surpassing human-level performance on imagenet classification[C]//Proceedings of IEEE International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2015:1026-1034.

选择文件类型/文献管理软件名称

选择包含的内容