浏览全部资源
扫码关注微信
纸质出版日期： 2023-09-16 ，
移动端阅览
曾庆旺， 董张玉， 杨学志， 种法亭. 2023. 结合全局上下文与融合注意力的干涉相位去噪. 中国图象图形学报， 28(09):2956-2968
Zeng Qingwang， Dong Zhangyu， Yang Xuezhi， Chong Fating. 2023. Interferometric phase denoising combining global context and fused attention. Journal of Image and Graphics， 28(09):2956-2968
目的
2
干涉相位去噪是合成孔径雷达干涉测量（interferometric synthetic aperture radar，InSAR）技术中的关键环节，其效果对测量精度具有重要影响。针对现有的干涉相位去噪方法大多关注局部特征以及在特征提取方面的局限性，同时为了平衡去噪和结构保持两者之间的关系，提出了一种结合全局上下文与融合注意力的相位去噪网络GCFA-PDNet（global context and fused attention phase denoising network）。
方法
2
将干涉相位分离为实部和虚部依次输入到网络，先从噪声相位中提取浅层特征，再将其映射到由全局上下文提取模块和融合注意力模块组成的特征增强模块，最后通过全局残差学习生成去噪图像。全局上下文提取模块能提取全局上下文信息，具有非局部方法的优势；融合注意力模块既强调关键特征，又能高效提取隐藏在复杂背景中的噪声信息。
结果
2
所提出的方法与对比方法中性能最优者相比，在模拟数据结果的平均峰值信噪比（peak signal to noise ratio， PSNR）和结构相似性（structural similarity， SSIM）指标分别提高了5.72%和2.94%，在真实数据结果的平均残差点减少百分比（percentage of residual point reduction， PRR）和相位标准偏差（phase standard deviation，PSD）指标分别提高了 2.01% 和3.57%。结合定性与定量分析，所提出的方法优于其他5种不同类型的相位去噪方法。
结论
2
提出的去噪网络较其他方法具有更强大的特征提取能力，此外由于关注全局上下文信息和强调关键特征，网络能够在增强去噪能力的同时保持原始相位细节。
Objective
2
Interferometric phase noise is introduced by three types of inherent factors： 1） system noise， such as thermal noise and synthetic aperture radar （SAR） speckle noise； 2） decoherence problems， including baseline， temporal， and spatial decoherence； 3） signal processing errors， such as misregistration. The existence of noise increases the difficulty of phase unwrapping and even causes the process to fail， thereby seriously interfering with the final interferometric result. Therefore， interferometric phase denoising is a key link in interferometric SAR （InSAR） technology. Its effect has an important influence on the accuracy of measurement results. The existing interferometric phase denoising algorithms still have many defects. First is the insufficient ability to capture global contextual information. Some algorithms ignore global context information or only focus on local context information derived from a few pixels. They also lack global context information. This feature is manifested as unstable detail preservation ability in denoising results. Second， many researchers only pay attention to the influence of the spatial dimension or channel dimension of the image on the denoising result to improve the performance of denoising networks. However， they do not use spatial and channel dimensions in combination. Third， the high-level features extracted from the deep layers of the convolutional neural network have rich semantic information and ambiguous spatial details. In comparison， the low-level features extracted from the shallow layers of the network contain considerable pixel-level noise information. However， these features are isolated from one another； thus， they cannot be fully used.
Method
2
Most of the existing interferometric phase denoising methods focus on local features， and they have many limitations in feature extraction. A phase denoising network called GCFA-PDNet is proposed to solve these problems while balancing the relationship between denoising and structure preservation. This proposed phase denoising network combines global context and fused attention. The method separates the interference phase into real and imaginary parts and inputs them into the network. First， the shallow features are extracted from the noise phase. Then， they are mapped to the feature enhancement module composed of the global context extraction module and the fused attention module. The shallow features extracted by the network are fused with the deep features concurrently. Finally， a denoised image is generated through global residual learning. Four global context extraction modules and four fused attention modules are used in the whole network. The core module of the global context extraction module is the global context block， which can extract the global context information and has the advantages of nonlocal methods. The fused attention module fuses the features extracted by its two submodules： the channel attention block and the spatial attention block. It emphasizes key features and efficiently extracts noise information hidden in complex backgrounds.
Result
2
We present six experimental results： Goldstein， nonlocal interferogram estimator （NL-InSAR）， block-matching 3D （InSAR-BM3D）， a deep learning framework for SAR interferometric phase restoration and coherence estimation （DeepInSAR）， phase filtering network， and GCFA-PDNet. Different evaluation indicators are selected for different datasets to evaluate the advantages and disadvantages of various algorithms objectively. For the experiments of simulated images， peak signal-to-noise ratio （PSNR） and structural similarity （SSIM） are selected as evaluation indicators. A large PSNR indicates a little difference between the filtered phase and the clean phase. However， it does not consider the correlation between pixels in the image. Therefore， SSIM is employed to assess the SSIM of the filtered images to evaluate the overall denoising quality of the denoised images. Compared with the comparative methods， the proposed method improves the average PSNR and SSIM indicators of the simulated data results by 5.72% and 2.94%， respectively. For the real interference phase， the above two indicators cannot be used for evaluating the denoising performance because no noise-free image is used as a reference. The number of residues （NOR） and phase standard deviation （PSD） can also be used as objective evaluation indicators to judge subjectively the denoising performance of real interferometric phase images. NOR can reflect the ability of the filtering method to suppress noise. The smaller the NOR of the filtered interference phase is， the stronger the ability to suppress noise is. PSD can be used to measure the discrete degree of noise distribution. The smaller the PSD value is， the more concentrated the noise distribution is and the better the quality of the interferogram is. In the real data results， compared with the comparative methods， the proposed method increases the average percentage of residual point reduction and PSD indicators by 2.01% and 3.57%， respectively. The visual observation of the experimental results of various algorithms shows that the proposed method achieves the best denoising results. The qualitative and quantitative analyses also indicate that the proposed method outperforms the five other types of phase denoising methods.
Conclusion
2
The phase denoising network designed in the present study combines the global context and fused attention. Thus， it has certain advantages compared with other related algorithms. It also has a more powerful feature extraction ability than other methods. The network focuses on global context information and emphasizes key features. Thus， it can maintain the original phase details while enhancing the denoising ability， thereby achieving the best denoising results.
合成孔径雷达干涉测量（InSAR）干涉相位去噪残差学习全局上下文融合注意力
synthetic aperture radar interferometry （InSAR）interferometric phase denoisingresidual learningglobal contextfused attention
Cao Y， Xu J R， Lin S， Wei F Y and Hu H. 2019. GCNet： non-local networks meet squeeze-excitation networks and beyond//Proceedings of 2019 IEEE/CVF International Conference on Computer Vision Workshop. Seoul， Korea （South）： IEEE： 1971-1980 ［DOI： 10.1109/ICCVW.2019.00246http://dx.doi.org/10.1109/ICCVW.2019.00246］
Deledalle C A， Denis L and Tupin F. 2011. NL-InSAR： nonlocal interferogram estimation. IEEE Transactions on Geoscience and Remote Sensing， 49（4）： 1441-1452 ［DOI： 10.1109/TGRS.2010.2076376http://dx.doi.org/10.1109/TGRS.2010.2076376］
Ferretti A， Monti-Guarnieri A V， Prati C M， Rocca F and Massonet D. 2007. InSAR principles： guidelines for SAR interferometry processing and interpretation ［EB/OL］. ［2022-05-20］. https://www.esa.int/esapub/tm/tm19/TM-19_ptA.pdfhttps://www.esa.int/esapub/tm/tm19/TM-19_ptA.pdf
Fu S H， Long X J， Yang X and Yu Q F. 2013. Directionally adaptive filter for synthetic aperture radar interferometric phase images. IEEE Transactions on Geoscience and Remote Sensing， 51（1）： 552-559 ［DOI： 10.1109/TGRS.2012.2202911http://dx.doi.org/10.1109/TGRS.2012.2202911］
Goldstein R M and Werner C L. 1998. Radar interferogram filtering for geophysical applications. Geophysical Research Letters， 25（21）： 4035-4038 ［DOI： 10.1029/1998GL900033http://dx.doi.org/10.1029/1998GL900033］
He K M， Zhang X Y， Ren S Q and Sun J. 2016. Deep residual learning for image recognition//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas， USA： IEEE： 770-778 ［DOI： 10.1109/cvpr.2016.90http://dx.doi.org/10.1109/cvpr.2016.90］
Hu J， Shen L and Sun G. 2018. Squeeze-and-excitation networks//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City， USA： IEEE： 7132-7141 ［DOI： 10.1109/CVPR.2018.00745http://dx.doi.org/10.1109/CVPR.2018.00745］
Lee J S， Papathanassiou K P， Ainsworth T L， Grunes M R and Reigber A. 1998. A new technique for noise filtering of SAR interferometric phase images. IEEE Transactions on Geoscience and Remote Sensing， 36（5）： 1456-1465 ［DOI： 10.1109/36.718849http://dx.doi.org/10.1109/36.718849］
Li Z W， Ding X L， Huang C， Jun Z J and Chen Y L. 2008. Improved filtering parameter determination for the Goldstein radar interferogram filter. ISPRS Journal of Photogrammetry and Remote Sensing， 63（6）： 621-634 ［DOI： 10.1016/j.isprsjprs.2008.03.001http://dx.doi.org/10.1016/j.isprsjprs.2008.03.001］
Liu S Q， Lei Y， Zhang L Y， Li B， Hu W M and Zhang Y D. 2022. MRDDANet： a multiscale residual dense dual attention network for SAR image denoising. IEEE Transactions on Geoscience and Remote Sensing， 60： #5214213 ［DOI： 10.1109/TGRS.2021.3106764http://dx.doi.org/10.1109/TGRS.2021.3106764］
Liu Y C， Shao Z R and Hoffmann N. 2021. Global attention mechanism： retain information to enhance channel-spatial interactions ［EB/OL］. ［2022-05-20］. http://arxiv.org/pdf/1512.03012.pdfhttp://arxiv.org/pdf/1512.03012.pdf
Lopez-Martinez C and Fabregas X. 2002. Modeling and reduction of SAR interferometric phase noise in the wavelet domain. IEEE Transactions on Geoscience and Remote Sensing， 40（12）： 2553-2566 ［DOI： 10.1109/TGRS.2002.806997http://dx.doi.org/10.1109/TGRS.2002.806997］
Lu Z H and Liu C. 2021. Multiscale feature reuse mixed attention network for image reconstruction. Journal of Image and Graphics， 26（11）： 2645-2658
卢正浩， 刘丛. 2021. 多尺度特征复用混合注意力网络的图像重建. 中国图象图形学报， 26（11）： 2645-2658 ［DOI： 10.11834/jig.200549http://dx.doi.org/10.11834/jig.200549］
Massonnet D and Feigl K L. 1998. Radar interferometry and its application to changes in the Earth’s surface. Reviews of Geophysics， 36（4）： 441-500 ［DOI： 10.1029/97RG03139http://dx.doi.org/10.1029/97RG03139］
Mukherjee S， Zimmer A， Sun X Y， Ghuman P and Cheng I. 2021. An unsupervised generative neural approach for InSAR phase filtering and coherence estimation. IEEE Geoscience and Remote Sensing Letters， 18（11）： 1971-1975 ［DOI： 10.1109/LGRS.2020.3010504http://dx.doi.org/10.1109/LGRS.2020.3010504］
Pu L M， Zhang X L， Zhou Z N， Shi J， Wei S J and Zhou Y Y. 2020. A phase filtering method with scale recurrent networks for InSAR. Remote Sensing， 12（20）： #3453 ［DOI： 10.3390/rs12203453http://dx.doi.org/10.3390/rs12203453］
Sica F， Cozzolino D， Zhu X X， Verdoliva L and Poggi G. 2018. InSAR-BM3D： a nonlocal filter for SAR interferometric phase restoration. IEEE Transactions on Geoscience and Remote Sensing， 56（6）： 3456-3467 ［DOI： 10.1109/TGRS.2018.2800087http://dx.doi.org/10.1109/TGRS.2018.2800087］
Sun X Y， Zimmer A， Mukherjee S， Kottayil N K， Ghuman P and Cheng I. 2020. DeepInSAR — A deep learning framework for SAR interferometric phase restoration and coherence estimation. Remote Sensing， 12（14）： #2340 ［DOI： 10.3390/rs12142340http://dx.doi.org/10.3390/rs12142340］
Suo Z Y， Zhang J Q， Li M， Zhang Q J and Fang C. 2016. Improved InSAR phase noise filter in frequency domain. IEEE Transactions on Geoscience and Remote Sensing， 54（2）： 1185-1195 ［DOI： 10.1109/TGRS.2015.2476355http://dx.doi.org/10.1109/TGRS.2015.2476355］
Tao L Q， Huang G M， Yang S C， Wang T T， Sheng H J and Fan H T. 2023. A interferogram denoising method based on convolutional neural network. Geomatics and Information Science of Wuhan University， 48（4）： 559-567
陶立清， 黄国满， 杨书成， 王童童， 盛辉军， 范海涛. 2023. 一种利用卷积神经网络的干涉图去噪方法. 武汉大学学报（信息科学版）， 48（4）： 559-567 ［DOI： 10.13203/j.whugis20200589http://dx.doi.org/10.13203/j.whugis20200589］
Wang Q L， Wu B G， Zhu P F， Li P H， Zuo W M and Hu Q H. 2020. ECA-Net： efficient channel attention for deep convolutional neural networks//Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）. Seattle， USA： IEEE： 11531-11539 ［DOI： 10.1109/CVPR42600.2020.01155http://dx.doi.org/10.1109/CVPR42600.2020.01155］
Wang X L， Girshick R， Gupta A and He K M. 2018. Non-local neural networks//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City， USA： IEEE： 7794-7803 ［DOI： 10.1109/CVPR.2018.00813http://dx.doi.org/10.1109/CVPR.2018.00813］
Wang Y， Huang H F， Dong Z and Wu M Q. 2016. Modified patch-based locally optimal Wiener method for interferometric SAR phase filtering. ISPRS Journal of Photogrammetry and Remote Sensing， 114： 10-23 ［DOI： 10.1016/j.isprsjprs.2016.01.013http://dx.doi.org/10.1016/j.isprsjprs.2016.01.013］
Xu G， Gao Y D， Li J W and Xing M D. 2020. InSAR phase denoising： a review of current technologies and future directions. IEEE Geoscience and Remote Sensing Magazine， 8（2）： 64-82 ［DOI： 10.1109/MGRS.2019.2955120http://dx.doi.org/10.1109/MGRS.2019.2955120］
Yang C Q， Zhou Y T， He H， Cui T and Wang Y. 2021. Global context and attention-based deep convolutional neural network for seismic data denoising. Geophysical Prospecting for Petroleum， 60（5）： 751-762， 855
杨翠倩， 周亚同， 何昊， 崔焘， 王杨. 2021. 基于全局上下文和注意力机制深度卷积神经网络的地震数据去噪. 石油物探， 60（5）： 751-762， 855 ［DOI： 10.3969/j.issn.1000-1441.2021.05.006http://dx.doi.org/10.3969/j.issn.1000-1441.2021.05.006］
相关作者
相关机构
京公网安备11010802024621