Cite this article
1
Download
54
Views
Journal Browser
Volume | Year
Issue
Search
News and Announcements
View All
ARTICLE

Self-guided attention denoising network for pre-stack seismic data: from coarse to fine

XINTONG DONG1,2 JUN LIN1,2 SHAOPING LU3,4 MING CHENG1,2 HONGZHOU WANG1,2
Show Less
1 College of Instrumentation and Electrical Engineering, Jilin University, Changchun 130026, P.R. China,
2 Southern Marine Science and Engineering Guangdong Laboratory (Zhanjiang),Zhanjiang 524000, P.R. China,
3 School of Earth Sciences and Engineering, Sun Yat-Sen University, Guangzhou 510275, P.R. China,
4 Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai 519000, P.R. China,
JSE 2023, 32(3), 271–300;
Submitted: 8 November 2022 | Accepted: 5 May 2023 | Published: 1 June 2023
© 2023 by the Authors. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution -Noncommercial 4.0 International License (CC-by the license) ( https://creativecommons.org/licenses/by-nc/4.0/ )
Abstract

Background noise attenuation is one of the most essential steps in seismic data processing. Residual background noise is likely to cause some artifacts in the following seismic imaging, thus bringing huge difficulties to the final interpretation. In recent years, deep-learning (DL) methods based on data driven strategy, especially the convolutional neural network (CNN), work well in seismic noise attenuation. In addition, it is applied automatically without parameter fine-tuning after training. To further improve their performance, we propose a novel architecture: self-guided attention network (SGA-Net) by combining self-guided strategy and spatial attention mechanism. Different from most of the conventional CNNs, this proposed SGA-Net can capture multi-scale features by performing the convolution operation on seismic data with different resolutions. In this network, the self-guided strategy is adopted to take full advantage of the multi-scale features; specifically, we utilize the global coarse features extracted at low resolution to guide the extraction process of local finer features at higher resolution. Furthermore, we design a spatial attention module with two inputs to fuse the global coarse and local fine features. We set up four competitive methods for SGA-Net including two traditional seismic denoising methods and two existing DL denoising methods in both synthetic and real experiments and experimental results demonstrate the advantage of SGA-Net both in noise attenuation and signal preservation.

Keywords
deep-learning
seismic noise attenuation
convolutional neural network
signal recovery
References
  1. Beckouche, S. and Ma, J., 2014. Simultaneous dictionary learning and denoising forseismic data. Geophysics, 79(3): A27-A31.
  2. Bekara, M. and van der Baan, M., 2007. Local singular value decomposition for signalenhancement of seismic data. Geophysics, 72(2): V59-V65.
  3. Birnie, C., Ravasi, M., Liu, S. and Alkhalifah, T., 2021. The potential of self-supervisednetworks for random noise suppression in seismic data. Artif. Intellig., 2: 47-59.
  4. Cadzow, J., 1988. Signal enhancement-a composite property mapping algorithm. IEEE
  5. Transact. Acoust., Speech Sign. Process., 36: 49-62.
  6. Chen, K. and Sacchi, M.D., 2015. Robust reduced-rank filtering for erratic seismic noiseattenuation. Geophysics, 80(1): V1-V11.
  7. Chen, K. and Sacchi, M.D., 2017. Robust f-x projection filtering for simultaneousrandom and erratic seismic noise attenuation. Geophys. Prosp., 65: 650-668.
  8. Chen, Y., 2020. Fast dictionary learning for noise attenuation of multidimensionalseismic data. Geophys. J. Internat., 222: 1717-1727.
  9. Chen, Y. and Ma, J., 2014. Random noise attenuation by fx empirical-modedecomposition predictive filtering. Geophysics, 79(3): V81-V91.
  10. Cheng, J., Chen, K. and Sacchi, M.D., 2015. Application of robust principal componentanalysis (RPCA) to suppress erratic noise in seismic records. Expanded Abstr.,4646-4651.
  11. Cui, Y., Xia, J., Wang, Z., Gao, S. and Wang, L., 2022. Lightweight spectral-spatialattention network for hyperspectral image classification. IEEE Transact. Geosci.Remote Sens., 60:1-14.
  12. Dong, X., Jiang, H., Zheng, S., Li, Y., and Yang, B., 2019a. Signal-to-noise ratioenhancement for 3C downhole microseismic data based on the 3D shearlet transformand improved back-propagation neural networks. Geophysics, 84(4): V245-V254.
  13. Dong, X., Li, Y. and Yang, B., 2019b. Desert low-frequency noise suppression by usingadaptive DnCNNs based on the determination of high-order statistic. Geophys. J.Internat., 219: 1281-1299.
  14. Dong, X. and Li, Y. 2021. Denoising the optical fiber seismic data by usingconvolutional adversarial network based on loss balance. IEEE Transact. Geosci.Remote Sens., 59: 10544-10554.
  15. Elboth, T., Presterud, I.V. and Hermansen, D., 2010. Time-frequency seismic datade-noising. Geophys. Prosp., 58: 441-453.
  16. Gomez, J.L., Velis, D.R. and Sabbione, J.I., 2020. Noise suppression in 2D and 3Dseismic data with data-driven sifting algorithms. Geophysics, 85(1): V1-V10.
  17. Gu, S., Guo, S., Zuo, W., Chen, Y., Timofte, R., Van Gool, L. and Zhang, L. 2020.
  18. Learned dynamic guidance for depth image reconstruction. IEEE Transact. PatternAnalys. Mach. Intellig., 42: 2437-2452.
  19. Gulunay, N., 1986. FX decon and complex Wiener prediction filter. Expanded Abstr.,90th Ann. Internat. SEG Mtg., Houston: 279-281.
  20. He, K., Zhang, X., Ren, S. and Sun, J., 2016. Deep residual learning for imagerecognition. IEEE Conf. Comput. Vis. Pattern Recognit., Las Vegas.
  21. Herrmann, F. and Hennenfent, G., 2008. Non-parametric seismic data recovery withcurvelet frames. Geophys. J. Internat., 173: 233-248.
  22. Hinton, G.E. and Salakhutdinov, R.R., 2006. Reducing the dimensionality of data withneural networks. Science, 313(5786): 504-507.
  23. Huang, N., Shen, Z., Long, S., Wu, M., Shih, H., Zheng, Q., Yen, N., Tung, C. and Liu,
  24. H., 1998. The empirical mode decomposition and Hilbert spectrum for nonlinear andnon-stationary time series analysis. Proc. Royal Soc. London, A 454: 903-995.
  25. Hui, T., Loy, C.C. and Tang, X., 2016. Depth map super-resolution by deep multi-scaleguidance. Europ. Conf. Comput. Vision, Amsterdam.
  26. Kaur, H., Fomel, S. and Pham, N., 2021. Seismic ground-roll noise attenuation usingdeep learning. Geophys. Prosp., 68: 2064-2077.
  27. Krohn, C., Ronen, S., Deere, J. and Gulunay, N., 2008. Introduction to this specialsection: Seismic noise. The Leading Edge, 27: 163-165.
  28. Lecun, Y., Bengio, Y. and Hinton, G.E., 2015. Deep learning. Nature, 521(7553):436-444.
  29. Liu, N., Li, F., Wang, D., Gao, J. and Xu, Z., 2022. Ground-roll separation andattenuation using curvelet-based multichannel variational mode decomposition. IEEETransact. Geosci. Remote Sens., 60: 1-14.
  30. Liu, W., Yan, Q. and Zhao, Y., 2020. Densely self-guided wavelet network for imagedenoising. IEEE Comput. Soc. Conf. Comput. Vision Pattern Recognit. Workshops,Seattle.
  31. Mousavi, S.M. and Langston, C.A., 2016. Hybrid seismic denoising using higher-orderstatistics and improved wavelet block thresholding. Bull. Seismol. Soc. Am., 106:1380-1393.
  32. Naghizadeh, M. and Sacchi, M.D., 2018. Ground-roll attenuation using curveletdownscaling. Geophysics, 83(3): V185-V195.
  33. Oropeza, V. and Sacchi, M.D., 2011. Simultaneous seismic data denoising andreconstruction via multichannel singular spectrum analysis. Geophysics, 76(3):V25-V32.
  34. Trickett, S., 2008. F-xy Cadzow noise suppression. CSPG CSEG CWLS Conv., LasVegas: 303-306.
  35. Saad, O.M. and Chen, Y., 2021. A fully unsupervised and highly generalized deeplearning approach for random noise suppression. Geophys. Prosp., 69: 709-726.
  36. Shan, H., Ma, J. and Yang H., 2009. Comparisons of wavelets, contourlets and curveletsin seismic denoising. J. Appl. Geophys., 69: 103-115.
  37. Simonyan, K. and Zisserman, A., 2015. Very deep convolutional networks forlarge-scale image recognition. 3rd Internat. Conf. Learning Representat., Montreal.
  38. Schneider, W.A., 1984. The common depth point stack. Proc. IEEE Conf., 72:1238-1254.
  39. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, L.and Polosukhin, I., 2017. Attention is all you need. Adv. Neural Informat. Process.Syst., 5998-6008.
  40. Wang, H., Chen, W., Huang, W., Zu, S., Liu, X., Yang, L. and Chen, Y., 2021.
  41. Nonstationary predictive filtering for seismic random noise suppression: a tutorial.Geophysics, 86(3): W21-W30.
  42. Wang, Z., Bovik, A.C., Sheikh, H.R. and Simoncelli, E.P., 2004. Image qualityassessment: from error visibility to structural similarity. IEEE Transact. ImageProcess., 13: 600-612.
  43. Wright, J., Peng, Y., Ma, Y., Ganesh, A. and Rao, S., 2009. Robust principal componentanalysis: Exact recovery of corrupted low-rank matrices by convex optimization. Adv.
  44. Neural Informat. Process. Syst., 22-Proc. 2009 Conf., Vancouver.
  45. Wu, X., Shi, Y., Fomel, S., Liang, L., Zhang, Q. and Yusifov, A., 2019. FaultNet3D:predicting fault probabilities, strikes, and dips with a single convolutional neuralnetwork. IEEE Transact. Geosci. Remote Sens., 57: 9138-9155.
  46. Wu, Z. and Huang, N.E., 2009. Ensemble empirical mode decomposition: Anoise-assisted data analysis method. Adv. Adapt. Data Analys., 1: 1-41.
  47. Yu, F. and Koltun, V., 2016. Multi-scale context aggregation by dilated convolutions.4th Internat. Conf. Learning Representat., San Juan, Puerto Rico.
  48. Yu, S. and Ma, J., 2021. Deep learning for Geophysics: current and future trends. Rev.Geophysics, 59(3), 1-36.
  49. Yu, S., Ma, J. and Wang, W., 2019. Deep learning for denoising. Geophysics, 84(6):V333-V350.
  50. Zhang, K., Zuo, W., Chen, Y., Meng, D. and Zhang, L., 2017. Beyond a Gaussiandenoiser: residual learning of deep CNN for image denoising. IEEE Transact. ImageProcess., 26: 3142-3155.
  51. Zhang, Z. and Alkhalifah, T., 2019. Regularized elastic full-waveform inversion usingdeep learning. Geophysics, 84(5): R741-R751.
  52. Zhao, H., Shi, J., Qi, X., Wang, X. and Jia, J., 2017. Pyramid scene parsing network.
  53. IEEE Conf. Computer Vision and Pattern Recognition, Honolulu.
  54. Zhong, T., Li, Y., Wu, N., Nie, P. and Yang, B., 2015. A study on the stationarity and
  55. Gaussianity of the background noise in land-seismic prospecting. Geophysics, 80(4):V67-V82.
  56. Zhu, W. and Beroza, G.C., 2019. PhaseNet: a deep-neural-network-based seismicarrival-time picking method. Geophys. J. Internat., 216: 261-273.
  57. Zhu, W., Mousavi, S.M. and Beroza, G.C., 2019. Seismic signal denoising anddecomposition using deep neural networks. IEEE Transact. Geosci. Remote Sens., 57:9476-9488.
Share
Back to top
Journal of Seismic Exploration, Electronic ISSN: 0963-0651 Print ISSN: 0963-0651, Published by AccScience Publishing