多级联注意力交互的沙尘图像增强方法

doi:10.11896/jsjkx.240800048

摘要/Abstract

摘要： 由于沙尘中悬浮粒子对光线的散射和吸收,计算机视觉系统采集到的图像往往存在颜色偏黄和对比度低的问题,目前退化图像增强算法大多针对去雾、去雨等情况,难以处理沙尘图像,针对沙尘图像增强有较大发展空间;由于缺少大规模数据集,神经网络难以实现对沙尘图像增强较好的鲁棒性。为此,提出一种基于多级联注意力交互的沙尘图像增强方法;此外,结合大气散射模型和深度信息构建了一个新的沙尘图像数据集。通过端到端的U-Net模型,提取多尺度特征图,使用多级联通道注意力交互模块融合多尺度特征图,使用多尺度卷积模块增强和恢复细节信息。实验结果表明,所提方法能有效去除图像中的沙尘和还原细节,在提出的数据集上优于先进方法的PSNR,SSIM和IPLPS指数。

关键词: 沙尘图像, 沙尘图像增强, 沙尘图像数据集, 特征融合, 注意力机制

Abstract: Due to the scattering and absorption of light by suspended particles,images captured in sand and dust conditions often suffer from a yellowish color bias and low contrast.Existing image enhancement algorithms for degraded images are mostly aimed at dehazing and deraining,making them difficult to effectively process sand dust images,thereby indicating a substantial potential for advancement in this field.The lack of large-scale datasets further complicates the task,hindering the ability of neural networks to robustly enhance sand dust images.To address this problem,a sand dust image enhancement method based on multi-cascaded attention interaction is proposed.In addition,we construct a new sand dust image dataset by combining the atmospheric scattering model and depth information.The method extracts multi-scale feature maps through an end-to-end U-Net model,fuses multi-scale feature maps using a multi-level channel attention interaction module,and enhances and restores detail information using a multi-scale convolution module.Experimental results show that the proposed method can effectively remove sand and dust in images and restore details,and achieves the best performance in terms of PSNR,SSIM,and IPLPS indices on the proposed dataset.

Key words: Sand dust images, Sand dust images enhancement, Sand dust images datasets, Feature fusion, Attention mechanism

中图分类号:

TP 391

王荣, 邹淑平, 郝鹏飞, 郭佳伟, 舒鹏. 多级联注意力交互的沙尘图像增强方法[J]. 计算机科学, 2025, 52(6A): 240800048-7. https://doi.org/10.11896/jsjkx.240800048

WANG Rong , ZOU Shuping, HAO Pengfei, GUO Jiawei, SHU Peng. Sand Dust Image Enhancement Method Based on Multi-cascaded Attention Interaction[J]. Computer Science, 2025, 52(6A): 240800048-7. https://doi.org/10.11896/jsjkx.240800048

参考文献

[1]CANTOR A.Optics of the atmosphere:scattering by mole-culesand particles[J].Optica Acta International Journal of Optics,1977,24(7):779-779.
[2]HE K,SUN J,TANG X.Single images haze removal using dark channel prior[J].IEEE Transactions on Pattern Analysis and Machine Intellience,2011,33(12):2341-2353.
[3]SULAMI M,GLATZER I,FATTAL R,et al.Automatic reco-very of the atmospheric light in hazy images[C]//2014 IEEE International Conference on Computational Photography(ICCP).Santa Clara,CA,USA,2014:1-11.
[4]PENG Y T,CAO K,COSMAN P C.Generalization of the Dark Channel Prior for Single Image Restoration[J].IEEE Transactions on Image Processing,2018,27(6):2856-2868.
[5]DOSOVITSKIY A,BEYER L,KOLESNIKOV A,et al.AnImage is Worth 16x16 Words:Transformers for Image Recognition at Scale[C]//2021 International Confer-ence on Learning Representations(ICLR).
[6]QIN X,WANG Z,BAI Y,et al.FFA-Net:Feature Fusion At-tention Network for Single Image Dehazing [C]// AAAI Conference on Artificial Intelligence.2020:11908-11915.
[7]ZHENG Y,ZHAN J,HE S,et al.Curricular Con-trastive Regularization for Physics-Aware Single Image Dehazing[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR).Vancouver,BC,Canada,2023:5785-5794.
[8]ZHENG Z.Ultra-High-Definition Image Dehazing via Multi-Guided Bilateral Learning[C]//2021 IEEE/CVF Con-ference on Computer Vision and Pattern Recognition(CVPR).Nashville,TN,USA,2021:16180-16189.
[9]DONG Y,LIU Y H,ZHANG H,et al.FD-GAN:Generative Adversarial Networks with Fusion-Discriminator for Single Image Dehazing[C]//AAAI Conference on Artificial Intelligence.34(7).
[10]VALANARASU J M J,YASARLA R,PATEL V M.TransWeather:Transformer-based Restoration of Images Degraded by Adverse Weather Conditions[C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR).New Orleans,LA,USA,2022:2343-2353.
[11]SONG Y,HE Z,QIAN H A,et al.Vision Transformers for Single Image Dehazing[J].IEEE Transactions on Image Proces-sing,2023,32:1927-1941.
[12]LIU Z.Swin Transformer:Hierarchical Vision Transformerusing Shifted Windows[C]//2021 IEEE/CVF International Conference on Computer Vision(ICCV).Montreal,QC,Canada,2021:9992-10002.
[13]PARK T H,EOM I K.Sand-Dust Image Enhancement Using Successive Color Balance With Coincident Chromatic Histogram[J].IEEE Access,2021,9:19749-19760.
[14]XU G,WANG X,XU X.Single image enhancement in sandstorm weather via tensor least square[J].IEEE/CAA Journal of Automatica Sinica,2020,7(6):1649-1661.
[15]FU X,HUANG Y,ZENG D,et al.A fusion-based enhancing approach for single sandstorm image[C]//2014 IEEE 16th International Workshop on Multi-media Signal Processing(MMSP).Jakarta,Indonesia,2014:1-5.
[16]SHI Z,FENG Y,ZHAO M,et al.Let You See in Sand DustWeather:A Method Based on Halo-Reduced Dark Channel Prior Dehazing for Sand-Dust Image Enhance-ment[J].IEEE Access,7:116722-116733.
[17]GAO G X,LAI H C,JIA Z H,et al.Sand-dust image restoration based on reversing the blue channel prior[J].IEEE Photonics Journal,2020,12(2):1-16.
[18]YU S Y,ZHU H,WANG J,et al.Single sand-dust image restoration using information loss constraint[J].Journal of Modern Optics,2016,63(21):2121-2130.
[19]MENG X L,HUANG J Y,LI Z Y,et al.DedustGAN:Unpaired learning for image dedusting based on Retinex with GANs[J].Expert Systems with Applications,2024,243:122844.
[20]LAND E H.The retinex theory of color vision[J].Scientific American,1977,237(6):108-128.
[21]ZHANG S L,TAO Z Y,LIN S.DedustNet:A Frequency dominated Swin Transformer based Wavelet Network for Agricultural Dust Removal[J].arXiv:2401.04750,2024.
[22]SANDLER M,HOWARD A,ZHU M,et al.MobileNetV2:Inverted Residuals and Linear Bottlenecks,[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.2018:4510-4520.
[23]HU J,SHEN L,SUN G.Squeeze and Excitation Networks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.2018:7132-7141.
[24]CHOLLET F.Xception:Deep Learning with Depthwise Separable Convolutions[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition(CVPR).2017:1800-1807.
[25]HE K,ZHANG X,REN S,et al.Deep residual learning forimage recognition[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR).2016:770-778.
[26]XIAN K.Monocular Relative Depth Perception with Web Stereo Data Supervision[C]//2018 IEEE/CVF Conference on Compu-ter Vision and Pattern Recognition.2018:311-320.
[27]ZHANG Y F,DING L,SHARMA G.HazeRD:An out-doorscene dataset and benchmark for single image dehazing[C]//2017 IEEE International Conference on Image Processing(ICIP).2017:3205-3209.
[28]WANG Z,BOVIK A C,SHEIKH H R,et al.Image quality assessment:from error visibility to structural similarity[J].IEEE Transactions on Image Processing,2004,13(4):600-612.
[29]ZHANG R,ISOLA P,EFROS A A,et al.The Unreasonable Effectiveness of Deep Features as a Perceptual Metric[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.2018:586-595.
[30]MITTAL A.Making a“completely blind”images quality analyzer[J].IEEE Signal Processing Letters,2013,20(3):209-212.
[31]FANG Y,MA K,WANG Z,et al.No-Reference Quality Assessment of Contrast-Distorted Images Based on Natural Scene Statistics[J].IEEE Signal Processing Letters,2015,22(7):838-842.
[32]LI B,PENG X,WANG Z,et al.AOD-Net:All-in-One Dehazing Network[C]//2017 IEEE International Conference on Compu-ter Vision(ICCV).2017:4780-4788.
[33]VENKATANATH N,PRANEETH D,CHANDRASEKHARB,et al.Blind images quality evaluation using perception based features[C]//2015 Twenty First National Conference on Communications(NCC).2015:1-6.
[34]WANG X,HAN X,HUANG W,et al.Multi-similarity loss with general pair weighting for deep metric learning[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR).USA,2019:5017-5025.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed