浏览全部资源
扫码关注微信
网络出版日期： 2024-10-16 ，
移动端阅览
周文俊,杨新龄,左承林等.MSFF-GAN：云雾环境下结冰风洞图像去雾模型[J].中国图象图形学报,
Zhou Wenjun,Yang Xinling,Zuo Chenglin,et al.MSFF-GAN： a Dehazing Model for Icing Wind Tunnel Images in Cloudy Conditions[J].Journal of Image and Graphics,
结冰风洞是地面试验的关键设备，可模拟云雾环境，对研究结冰对飞机性能的影响极为重要。但云雾环境会降低图像质量，这不仅阻碍了对结冰过程的细致观察，还减少了结冰检测与分析的准确度。本文提出了一种新的图像去雾方法——多尺度特征融合生成对抗网络（multi-scale feature fusion generative adversarial network， MSFF-GAN），旨在改善结冰风洞云雾环境下的图像质量，提高研究精度。通过利用生成对抗网络的能力，MSFF-GAN高效去除结冰风洞图像的雾，核心在于其生成器的特征融合和增强策略。特征融合模块通过反投影技术精准融合了图像多尺度特征，增强策略模块通过简洁网络结构细化中间结果，优化图像质量。本文还设计了一种先验特征融合模块，有效整合至网络中。此外，通过多尺度判别器策略获得全面上下文信息，显著提升视觉质量。同时，采用多重损失函数共同优化去雾模型，确保最优去雾效果。在六种结冰风洞云雾场景下，对比实验了本文提出的MSFF-GAN去雾方法与其他传统及深度学习方法。实验结果显示，结冰风洞云雾环境下MSFF-GAN生成的去雾图像更清晰，去雾效果更显著，且在相关评价指标上表现优异。MSFF-GAN在结冰风洞云雾环境中展示出卓越的去雾效果和良好的泛化性，为结冰风洞图像的清晰化处理提供了新思路，有望为飞机结冰与防除冰研究提供更精准、可靠的机翼结冰图像数据。
Abstract： Objective During high-altitude flights， aircraft are often exposed to extremely low-temperature environments， particularly below the freezing point. Such conditions make it easy for water vapor and liquid water in contact with the aircraft surface to condense and form ice layers. This icing phenomenon poses a significant threat to flight safety as it can alter the aircraft's surface structure， affecting its aerodynamic performance and potentially leading to serious flight accidents. Therefore， the study of aircraft icing is crucial， as it relates to the safeguarding of flight safety and the enhancement of aircraft performance. To delve deeper into the impact of aircraft icing on flight performance， icing wind tunnels play a pivotal role as ground test facilities. They are capable of simulating high-altitude cloud and hazy environments， enabling researchers to replicate the icing conditions that aircraft may encounter during flight and observe and analyze the effects of icing on aircraft performance. However， in cloudy and haze environments， the quality of captured images is often severely compromised. Atmospheric scattering effects can cause images to appear grayish， reduce contrast， and obscure the originally visible ice structures. This not only hinders researchers' observations and documentation of the icing process but also redu
ces the accuracy of icing detection， thereby affecting subsequent research and analysis. Therefore， preprocessing the captured images before utilizing the icing wind tunnel for aircraft icing research is particularly important. Image dehazing techniques， as effective image processing methods， can significantly enhance image quality， making originally blurred ice structures clearly visible. By applying dehazing techniques， not only can researchers improve their observation of the icing process， but also enhance the accuracy of icing detection， providing more reliable data support for subsequent research and analysis. Method Traditional image dehazing methods suffer from issues such as parameter sensitivity and long processing time， which directly impact the effectiveness and efficiency of dehazing. In recent years， image dehazing methods based on deep learning have garnered widespread research attention. Through end-to-end learning， this approach demonstrates strong adaptability and dehazing capabilities. However， deep learning-based dehazing methods require a significant amount of labeled data and high computational resources for support. To effectively address the issues of insufficient feature extraction and residual haze in current deep learning-based image dehazing methods when processing icing wind tunnel images， this paper proposes a generative adversarial network with multi-scale feature fusion for image dehazing （MSFF-GAN）. MSFF-GAN aims to leverage the exceptional image generation capabilities and supervised learning characteristics of GANs to achieve more automated and efficient dehazing. First of all， the generative adversarial network introduces a competitive mechanism to make the generator and the discriminator compete with each other in the training process， thus continuously optimizing the generation results. This mechanism helps the generator learn more complex image distributions and generate images with higher quality. Secondly， the generative adversarial network adopts supervised learning for t
raining， which means that it can directly use a large amount of labeled data for learning， so as to better understand the inherent structure and characteristics of images. The generative adversarial network consists of a generator and a discriminator. The generator primarily consists of two modules： feature fusion and enhancement strategies. The feature fusion module effectively integrates multi-scale features of the image by employing back-projection techniques. The enhancement strategy module， on the other hand， achieves gradual refinement of intermediate results through a concise network design， thereby enhancing the overall image quality. To extract richer image information from hazy images and further enhance the dehazing effect， we extract prior features of hazy images： dark channel and color attenuation， and integrate them into the network structure. Furthermore， by setting the discriminator in the GAN-based dehazing network to be multi-scale， it can synthesize information from different scales， providing more comprehensive contextual information and feedback. This approach improves the visual quality of the image across multiple receptive fields. Finally， to further enhance the dehazing effect， multiple loss functions are employed to jointly constrain the training of the dehazing model. Result We selects six different icing and haze images of aircraft wings in various cloud and haze scenarios within the wind tunnel， under conditions of different observation angles and with the values of MVD and LWC set to 25
<math id="M1"><mi>μ</mi><mi>m</mi></math>
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492499&type=
2.87866688
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492494&type=
4.31799984
and 1.31
<math id="M2"><mi>g</mi><mo>/</mo><msup><mrow><mi>m</mi></mrow><mrow><mn mathvariant="normal">3</mn></mrow></msup></math>
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492478&type=
3.21733332
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492479&type=
5.84200001
， 20
<math id="M3"><mi>μ</mi><mi>m</mi></math>
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492499&type=
2.87866688
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492494&type=
4.31799984
and 1.0
<math id="M4"><mi>g</mi><mo>/</mo><msup><mrow><mi>m</mi></mrow><mrow><mn mathvariant="normal">3</mn></mrow></msup></math>
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492482&type=
3.21733332
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492498&type=
5.84200001
， and 20
<math id="M5"><mi>μ</mi><mi>m</mi></math>
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492499&type=
2.87866688
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492494&type=
4.31799984
and 0.5
<math id="M6"><mi>g</mi><mo>/</mo><msup><mrow><mi>m</mi></mrow><mrow><mn mathvariant="normal">3</mn></mrow></msup></math>
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492495&type=
3.21733332
https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=67492496&type=
5.84200001
， respectively， to form the test sets. Subsequently， experiments are conducted on these six different icing wind tunnel scene test sets， comparing the method proposed in this paper with four traditional dehazing methods and four deep learning-based dehazing methods. The experimental results demonstrate that the dehazing images obtained by our method are the clearest and achieve the best dehazing effect. Notably， our method produces dehazed images with higher clarity， effectively preserving the original colors and texture information of the icing wind tunnel images， especially without compromising the shape of the icing regions on the wings. Meanwhile， we conducted experiments on synthetic datasets from public datasets as well as real foggy images， and the results demonstrated significant defogging effects， resulting in an improvement in image quality. Furthermore， we conducted ablation experiments to verify the effectiveness of our proposed dehazing method. These experiments confirmed that our method significantly improves dehazing performance and achieves better
results in evaluation metrics. These positive outcomes highlight the robustness and generalization capabilities of our dehazing model in various cloud and haze environments within the wind tunnel. Conclusion The dehazing model proposed in this paper has demonstrated satisfactory dehazing effects and excellent generalization performance in icing wind tunnel simulations with varying degrees of haze concentration. The model is capable of effectively eliminating the haze from the images， significantly restoring their clarity， and providing researchers with improved visual outcomes. Additionally， the model preserves the shape and color of the icing region on the aircraft wing to a large extent while also reconstructing crucial parts of the background area around the wing. This provides researchers with clearer image information， which is crucial for subsequent icing detection and related work.
结冰风洞云雾环境机翼结冰图像去雾生成对抗网络多尺度特征融合
icy wind tunnelcloud environmentwing icing image dehazinggenerating adversarial networksmulti-scale feature fusion
Cui Y and Knoll A. 2023. Exploring the potential of channel interactions for image restoration. Knowledge-Based Systems， 282： 111156 ［DOI： 10.1016/j.knosys.2023.111156http://dx.doi.org/10.1016/j.knosys.2023.111156］
Dai S Y， Han M， Wu Y and Gong Y H. 2007. Bilateral back-projection for single image super resolution//2007 IEEE International Conference on Multimedia and Expo. Beijing， China： IEEE： 1039-1042 ［DOI： 10.1109/ICME.2007.4284831http://dx.doi.org/10.1109/ICME.2007.4284831］
Deniz E， Anil G and Hazim K E. 2018. Cycle-Dehaze： Enhanced CycleGAN for Single Image Dehazing//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition （CVPR） Workshops. Salt Lake City， UT， USA. IEEE： 825-833 ［DOI： 10.1109/CVPRW.2018.00127http://dx.doi.org/10.1109/CVPRW.2018.00127］
Dong H， Pan J， Xiang L， Hu Z， Zhang X， Wang F and Yang M H. 2020. Multi-scale boosted dehazing network with dense feature fusion//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. Seattle， WA， USA： IEEE： 2157-2167 ［DOI： 10.1109/CVPR42600.2020.00223http://dx.doi.org/10.1109/CVPR42600.2020.00223］
Dong Y， Liu Y H， Zhang H， Chen S F and Qiao Y. 2020. FD-GAN： Generative Adversarial Networks with Fusion-Discriminator for Single Image Dehazing//Proceedings of the AAAI Conference on Artificial Intelligence. New York， USA. AAAI Press： 34（07）： 10729-10736 ［DOI： 10.1609/aaai.v34i07.6701http://dx.doi.org/10.1609/aaai.v34i07.6701］
Guo X D， Zhang P T， Zhao X L， Yang S K and Lin W. 2020. The compliance verification of thermodynamic flowfield in the large icing wind tunnel. Journal of Experiments in Fluid Mechanics， 34（5）： 79-88
郭向东， 张平涛， 赵献礼， 杨升科， 林伟. 2020. 大型结冰风洞热流场符合性验证. 实验流体力学， 34（5）： 79-88 ［DOI： 10.11729/syltlx20190113http://dx.doi.org/10.11729/syltlx20190113］
Guo L， Cheng Y and Wang Z X. 2018. Experimental study on droplet size measurement and control of icing cloud in icing wind tunnel. Journal of Experiments in Fluid Mechanics， 32（2）： 55-60
郭龙， 程尧， 王梓旭. 2018. 结冰风洞试验段云雾粒径测量与控制实验研究. 实验流体力学， 32（2）： 55-60 ［DOI： 10.11729/syltlx20170096http://dx.doi.org/10.11729/syltlx20170096］
Goodfellow I， Pouget-Abadie J， Mirza M， Xu B， Warde-Farley D， Ozair S， Courville A and Bengio Y. 2014. Generative adversarial nets. Advances in neural information processing systems， 27： ［DOI： 10.1145/3422622］
Galdran A. 2018. Image dehazing by artificial multiple-exposure image fusion. Signal Processing， 149： 135-147 ［DOI： 10.1016/j.sigpro.2018.03.008http://dx.doi.org/10.1016/j.sigpro.2018.03.008］
Guo M， Xiong F， Zhao B， Huang Y， Xie Z， Wu L， Chen X， Zhang J. 2024. TDEGAN： A Texture-Detail-Enhanced Dense Generative Adversarial Network for Remote Sensing Image Super-Resolution. Remote Sensing， 16（13）： 2312. ［DOI： 10.3390/rs16132312http://dx.doi.org/10.3390/rs16132312］
He K， Sun J and Tang X. 2010. Single image haze removal using dark channel prior. IEEE transactions on pattern analysis and machine intelligence， 33（12）： 2341-2353 ［DOI： 10.1109/TPAMI.2010.168http://dx.doi.org/10.1109/TPAMI.2010.168］
Hautiere N， Tarel J P， Aubert D and Dumont E. 2008. IBlind contrast enhancement assessment by gradient ratioing at visible edges. Image Analysis & Stereology， 27（2）： 87-95 ［DOI： 10.5566/ias.v27.p87-95http://dx.doi.org/10.5566/ias.v27.p87-95］
Huang Y，Peng H，Li C S，Gao S M and Chen F. 2024. LLFlowGAN： a low-light image enhancement method for constraining invertible flow in a generative adversarial manner. Journal of Image and Graphics，29（01）：0065-0079
黄颖，彭慧，李昌盛，高胜美，陈奉 . 2024. LLFlowGAN：以生成对抗方式约束可逆流的低照度图像增强. 中国图象图形学报，29（01）：0065-0079［DOI： 10. 11834/jig. 230063http://dx.doi.org/10.11834/jig.230063］
Isola P， Zhu J Y， Zhou T and Efros A. 2017. Image-to-image translation with conditional adversarial networks//Proceedings of the IEEE conference on computer vision and pattern recognition. Honolulu， HI， USA： IEEE： 1125-1134 ［DOI： 10.1109/CVPR.2017.632http://dx.doi.org/10.1109/CVPR.2017.632］
Johnson J， Alahia and Fei-fei L. 2016. Perceptual losses for real-time style transfer and super-resolution//Computer Vision–ECCV 2016： 14th European Conference. Amsterdam， The Netherlands： Springer： 694-711 ［DOI： 10.1007/978-3-319-46475-6_43http://dx.doi.org/10.1007/978-3-319-46475-6_43］
Liu S Y， Wang Q， Yi X， Zhang P T， Zuo C L and Guo Q L. 2023. New progress of 3m x 2m icing wind tunnel test technology from 2020 to 2022. Journal of Experiments in Fluid Mechanics， 41（1）： 57-65
刘森云， 王桥， 易贤， 张平涛， 左承林， 郭奇灵. 2023. 3m×2m结冰风洞试验技术新进展 （2020- 2022 年）. 空气动力学学报， 41（1）： 57-65 ［DOI： 10.7638/kqdlxxb-2022.0167http://dx.doi.org/10.7638/kqdlxxb-2022.0167］
Luo J， Chang T and Bo Q. 2022. Multi-feature Fusion Network for Single Image Dehazing//Pattern Recognition and Computer Vision. Shenzhen， China： Springer： 127-136 ［DOI： 10.1007/978-3-031-18916-6_11http://dx.doi.org/10.1007/978-3-031-18916-6_11］
Li B Y， Peng X L， Wang Z Y， Xu J Z and Feng D. 2017. Aod-net： All-in-one dehazing network//Proceedings of the IEEE international conference on computer vision. Venice， Italy： IEEE： 4770-4778 ［DOI： 10.1109/ICCV.2017.511http://dx.doi.org/10.1109/ICCV.2017.511］
Lu L P， Xiong Q， Chu D F and Xu B R. 2023. MixDehazeNet： Mix Structure Block For Image Dehazing Network. arxiv preprint arxiv：2305.17654 ［DOI： 10.48550/arXiv.2305.17654http://dx.doi.org/10.48550/arXiv.2305.17654］
Lan Z，Yan C P，Li H and Zheng Y D. 2023. HDA-GAN：hybrid dual attention generative adversarial network for image inpainting. Journal of Image and Graphics，28（11）：3440-3452
兰治，严彩萍，李红，郑雅丹. 2023. 混合双注意力机制生成对抗网络的图像修复模型. 中国图象图形学报，28（11）：3440-3452［DOI： 10. 11834/jig. 220919http://dx.doi.org/10.11834/jig.220919］
Mittal A， Soundararajan R and Bovik A C. 2012. Making a “completely blind” image quality analyzer. IEEE Signal processing letters， 20（3）： 209-212 ［DOI： 10.1109/LSP.2012.2227726http://dx.doi.org/10.1109/LSP.2012.2227726］
Qin X， Wang Z L， Bai Y D， Xie X D and Jia H Z. 2020. FFA-Net： Feature fusion attention network for single image dehazing//Proceedings of the AAAI Conference on Artificial Intelligence. New York， USA： AAAI Press： 11908-11915 ［DOI： 10.1609/aaai.v34i07.6865http://dx.doi.org/10.1609/aaai.v34i07.6865］
Qu Y， Chen Y， Huang J and Xie Y. 2019. Enhanced Pix2pix Dehazing Network//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）. Long Beach， CA， USA. IEEE： 8160-8168. ［DOI： 10.1109/CVPR.2019.00835http://dx.doi.org/10.1109/CVPR.2019.00835］
Ronneberger O， Fischer P and Brox T. 2015. U-net： Convolutional networks for biomedical image segmentation//International Conference on Medical image computing and computer-assisted intervention. Munich， Germany： Springer： 234-241 ［DOI： 10.1007/978-3-319-24574-4_28http://dx.doi.org/10.1007/978-3-319-24574-4_28］
Romano Y and Elad M. 2015. Boosting of image denoising algorithms. SIAM Journal on Imaging Sciences， 8（2）： 1187-1219 ［DOI： 10.1137/140990978http://dx.doi.org/10.1137/140990978］
Raikwar S C and Tapaswi S. 2020. Lower bound on transmission using nonlinear bounding function in single image dehazing. IEEE Transactions on Image Processing， 29： 4832-4847 ［DOI： 10.1109/TIP.2020.2975909http://dx.doi.org/10.1109/TIP.2020.2975909］
Ren W Q， Liu S， Zhang H， Pan J S， Cao X C and Yang M H. 2016. Single image dehazing via multi-scale convolutional neural networks//Computer Vision–ECCV 2016. Amsterdam， The Netherlands： Springer： 154-169 ［DOI： 10.1007/978-3-319-46475-6_10http://dx.doi.org/10.1007/978-3-319-46475-6_10］
Tu H， Wang W， Chen J， Li G and Wu F. 2022. Dehazing algorithm combined with atmospheric scattering model based on generative adversarial network. Journal of Zhejiang University： Engineering Science， 56（02）：225-235
屠杭垚， 王万良， 陈嘉诚， 李国庆， 吴菲. 2022. 结合大气散射模型的生成对抗网络去雾算法. 浙江大学学报（工学版）. 56（02）：225-235. ［DOI： 10.3785/j.issn.1008-973X.2022.02.002http://dx.doi.org/10.3785/j.issn.1008-973X.2022.02.002］
Wang D W， Li S L， Han P F and Ren X C. 2021. Feature Constraint CycleGAN for Single image Dehazing Algorithm. Laser & Optoelectronics Progress， 58（14）： 1410017
王殿伟， 李顺利， 韩鹏飞， 刘颖， 姜静， 任新成. 2021. 基于特征约束 CycleGAN 的单幅图像去雾算法研究. 中国激光杂志社， 58（14）： 1410017 ［DOI： 10.3788/LOP202158.1410017http://dx.doi.org/10.3788/LOP202158.1410017］
Wang R and Yang Y. 2023. Multi-level fusion dehazing network based on learning of hazy layer. Journal of Measurement Science & Instrumentation， 14（2）： 200. ［DOI： 10.62756/jmsi.1674-8042.2023023http://dx.doi.org/10.62756/jmsi.1674-8042.2023023］
Wu H Y， Qu Y Y， Lin S H， Zhou J， Qiao R Z， Zhang Z Z， Xie Y and Ma L Z. 2021. Contrastive learning for compact single image dehazing//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Nashville， TN， USA： IEEE： 10551-10560 ［DOI： 10.1109/CVPR46437.2021.01041http://dx.doi.org/10.1109/CVPR46437.2021.01041］
Wang， S， Mei， X， Kang， P. et al. 2024. DFC-dehaze： an improved cycle-consistent generative adversarial network for unpaired image dehazing. The Visual Computer 40， 2807–2818 ［DOI： 10.1007/s00371-023-02987-8http://dx.doi.org/10.1007/s00371-023-02987-8］
Yi X， Wang B， Li W B and Guo L. 2017. Research progress on ice shape measurement approaches for aircraft icing. Journal of Experiments in Fluid Mechanics， 38（2）： 13-24
易贤， 王斌， 李伟斌， 郭龙. 2017. 飞机结冰冰形测量方法研究进展. 航空学报， 38（2）： 13-24 ［DOI： 10.7527/S1000-6893.2016.0276http://dx.doi.org/10.7527/S1000-6893.2016.0276］
Yang S K， Ni Z S， Xiao C H， et al. 2020. Research on application of image dehazing algorithm in icing wind tunnel. Transducer and Microsystem Technologies， 39（12）： 50-52
杨升科， 倪章松， 肖春华， 王茂. 2020. 图像去雾算法在结冰风洞中的应用研究. 传感器与微系统， 39（12）： 50-52 ［DOI： 10.11729/syltlx20170096http://dx.doi.org/10.11729/syltlx20170096］
Yang Y， Wang C Y， Liu R S， Zhang L， Guo X J and Tao D C. 2022. Self-Augmented Unpaired Image Dehazing via Density and Depth Decomposition//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans， LA， USA： New Orleans， LA， USA： 2037-2046 ［DOI： 10.1109/CVPR52688.2022.00208http://dx.doi.org/10.1109/CVPR52688.2022.00208］
Zheng Y， Su J， Zhang S， Tao M and Wang L. 2022. Dehaze-AGGAN： Unpaired Remote Sensing Image Dehazing Using Enhanced Attention-Guide Generative Adversarial Networks. IEEE Transactions on Geoscience and Remote Sensing， 60. ［DOI： 10.1109/TGRS.2022.3204890http://dx.doi.org/10.1109/TGRS.2022.3204890］
Zhu Q S， Mai J M and Shao L. 2015. A fast single image haze removal algorithm using color attenuation prior. IEEE transactions on image processing， 24（11）： 3522-3533 ［DOI： 10.1109/TIP.2015.2446191http://dx.doi.org/10.1109/TIP.2015.2446191］
Zamir S W， Arora A， Khan S， Hayat M， Khan F S， Yang M H and Shao L. 2021. Multi-stage progressive image restoration//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Nashville， TN， USA： IEEE： 14821-14831 ［DOI： 10.1109/CVPR46437.2021.01458http://dx.doi.org/10.1109/CVPR46437.2021.01458］
Zhang M X，Wei R，Liu B，Xu S P，Bai X Z and Zhou F G. 2023. CG-DRR：digital reconstructed radiograph generation algorithm based on Cycle-GAN. Journal of Image and Graphics，28（04）：1212-1222
张孟希，魏然，刘博，徐寿平，白相志，周付根 . 2023. CG-DRR：数字重建放射影像循环一致性生成对抗算法. 中国图象图形学报，28（04）：1212-1222［DOI： 10. 11834/jig. 210868http://dx.doi.org/10.11834/jig.210868］
相关作者
相关机构
京公网安备11010802024621