Unsupervised seismic data deblending based on the convolutional autoencoder regularization

• Autorzy: Xue Yaru , Chen Yuyao , Jiang Minhui , Duan Hanting , Niu Libo , Chen Chong

Identyfikatory

DOI

10.1007/s11600-022-00772-0

• Języki publikacji: EN

Abstrakty

EN

Simultaneous source technology can provide high-quality seismic data with lower acquisition costs. However, a deblending algorithm is needed to suppress the blending noise. The supervised deep learning methods are effective, but are usually limited by the lack of labels. To solve the problem, we propose an unsupervised deep learning method based on acquisition system. A convolutional autoencoder (CAE) network is employed to predict the deblending results of the input pseudodeblended data. And then, the deblending results will be re-blended using the given blending operator. The parameters of CAE will be optimized by the difference between re-blended data and input data, which is defined as ‘blending loss.’ The blending problem is ill-posed but the CAE can be regarded as an implicit regularization term which constrains the solving process to obtain the desire solution. A numerical test on synthetic data demonstrates that the proposed method can converge to correct results and two field data experiments verify the flexibility and effectiveness of our model. The transfer training method is also used to improve model performance.

Słowa kluczowe

EN

simultaneous source data; unsupervised learning; convolutional autoencoder network; deblending

PL

jednoczesne dane źródłowe; uczenie bez nadzoru

• Wydawca: Instytut Geofizyki PAN ; Springer
• Czasopismo: Acta Geophysica
• Rocznik: 2022
• Tom: Vol. 70, no 3
• Strony: 1171--1182
• Opis fizyczny: Bibliogr. 32 poz.

Twórcy

autor

Xue Yaru

• State Key Laboratory of Petroleum Resources and Prospecting, College of Information Science and Engineering, China University of Petroleum (Beijing), Beijing 102249, China

autor

Chen Yuyao

• State Key Laboratory of Petroleum Resources and Prospecting, College of Information Science and Engineering, China University of Petroleum (Beijing), Beijing 102249, China

autor

Jiang Minhui

• State Key Laboratory of Petroleum Resources and Prospecting, College of Information Science and Engineering, China University of Petroleum (Beijing), Beijing 102249, China

autor

Duan Hanting

• State Key Laboratory of Petroleum Resources and Prospecting, College of Information Science and Engineering, China University of Petroleum (Beijing), Beijing 102249, China

autor

Niu Libo

• State Key Laboratory of Petroleum Resources and Prospecting, College of Information Science and Engineering, China University of Petroleum (Beijing), Beijing 102249, China

autor

Chen Chong

• State Key Laboratory of Petroleum Resources and Prospecting, College of Information Science and Engineering, China University of Petroleum (Beijing), Beijing 102249, China

Bibliografia

• 1. Abma RL, Yan J (2009) Separating simultaneous sources by inversion. 71st EAGE conference and exhibition incorporating SPE EUROPEC 2009 https://doi.org/10.3997/2214-4609.201400403.
• 2. Akerberg P, Hampson G, Rickett J, Martin H, Cole J (2008) Simultaneous source separation by sparse Radon transform. SEG Tech Prog Expand Abstr 2008:2801–2805. https://doi.org/10.1190/1.3063927
• 3. Bagaini C, Daly M, Moore I (2012) The acquisition and processing of dithered slip-sweep vibroseis data. Geophys Prospect 60(4):618–639. https://doi.org/10.1111/j.1365-2478.2012.01085.x
• 4. Beckouche S, Ma J (2014) Simultaneous dictionary learning and denoising for seismic data. Geophysics 79(3):A27–A31. https://doi.org/10.1190/geo2013-0382.1
• 5. Berkhout AJG, Blacquière G, Verschuur E (2008) From simultaneous shooting to blended acquisition. SEG Tech Progr Expand Abstr 2008:2831–2838. https://doi.org/10.1190/1.3063933
• 6. Chen Y, Fomel S, Hu J (2014) Iterative deblending of simultaneous-source seismic data using seislet-domain shaping regularization. Geophysics 79(5):V179–V189. https://doi.org/10.1190/geo2013-0449.1
• 7. Cheng J, Sacchi MD (2015) Separation and reconstruction of simultaneous source data via iterative rank reduction. Geophysics 80(4):V57–V66. https://doi.org/10.1190/geo2014-0385.1
• 8. Cheng J, Sacchi M, Gao J (2018) Computational efficient multidimensional singular spectrum analysis for prestack seismic data reconstruction. Geophysics 84(2):V111–V119. https://doi.org/10.1190/geo2018-0343.1
• 9. Doulgeris P, Mahdad A, Blacquière G (2010) Separation of blended data by iterative estimation and subtraction of interference noise. SEG Tech Progr Expand Abstr 2010:3514–3518. https://doi.org/10.1190/1.3513579
• 10. Dragoset W et al (2009) A 3D wide-azimuth field test with simultaneous marine sources. 71st EAGE conference and exhibition incorporating SPE EUROPEC 2009 https://doi.org/10.3997/2214-4609.201400570.
• 11. Fomel S (2007) Local seismic attributes. Geophysics 72(3):A29–A33. https://doi.org/10.1190/1.2437573
• 12. Gan S, Wang S, Chen Y, Chen X (2016) Simultaneous-source separation using iterative seislet-frame thresholding. IEEE Geosci Remote Sens Lett 13(2):197–201. https://doi.org/10.1109/LGRS.2015.2505319
• 13. Hampson G, Stefani J, Herkenhoff F (2008) Acquisition using simultaneous sources. Lead Edge 27(7):918–923. https://doi.org/10.1190/1.2954034
• 14. Huo S, Luo Y, Kelamis P (2009) Simultaneous sources separation via multi-directional vector-median filter. SEG Tech Progr Expand Abstr 2009:31–35. https://doi.org/10.1190/1.3255522
• 15. Ibrahim A, Sacchi MD (2013) Simultaneous source separation using a robust Radon transform. Geophysics 79(1):V1–V11. https://doi.org/10.1190/geo2013-0168.1
• 16. Ioffe S, Szegedy C (2015) Batch normalization: accelerating deep network training by reducing internal covariate shift. arXiv e-prints: arXiv:1502.03167
• 17. Kingma DP, Ba J (2014) Adam: a method for stochastic optimization. arXiv e-prints: arXiv:1412.6980
• 18. Mahdad A (2012) Deblending of seismic data. The Netherlands. https://doi.org/10.4233/uuid:68883d84-3cf3-4b5c-a8ee-9c0a92604fce
• 19. Matharu G et al (2020) Simultaneous source deblending using a deep residual network. SEG 2019 workshop: mathematical geophysics: traditional vs learning, Beijing, China, 5–7 November 2019: 13–16. https://doi.org/10.1190/iwmg2019_04.1
• 20. Moore I et al (2008) Simultaneous source separation using dithered sources. SEG Tech Progr Expand Abstr 2008:2806–2810. https://doi.org/10.1190/1.3063928
• 21. Sun J, Slang S, Elboth T, Greiner TL, McDonald S, Gelius LJ (2020) A convolutional neural network approach to deblending seismic data. Geophysics 85(4):WA13–WA26. https://doi.org/10.1190/geo2019-0173.1
• 22. Tukey J (1974) Nonlinear (nonsuperposal) methods for smoothing data. Congr Rec Eascon 673
• 23. Xue Y, Man M, Zu S, Chang F, Chen Y (2017) Amplitude-preserving iterative deblending of simultaneous source seismic data using high-order Radon transform. J Appl Geophys 139:79–90. https://doi.org/10.1016/j.jappgeo.2017.02.010
• 24. Xue Y, Niu L, Chen C, Xu X (2021) An adaptive-rank singular spectrum analysis for simultaneous-source data separation. IEEE Geosci Remote Sens Lett 18(5):801–805. https://doi.org/10.1109/LGRS.2020.2989750
• 25. Yu S, Ma J, Zhang X, Sacchi MD (2015) Interpolation and denoising of high-dimensional seismic data by learning a tight frame. Geophysics 80(5):V119–V132. https://doi.org/10.1190/geo2014-0396.1
• 26. Zhang M, Liu Y, Chen Y (2019) Unsupervised seismic random noise attenuation based on deep convolutional neural network. IEEE Access 7:179810–179822. https://doi.org/10.1109/ACCESS.2019.2959238
• 27. Zhou Y, Chen W, Gao J, Pascal F (2013) Seismic deblending by sparse inversion over dictionary learning. SEG Tech Progr Expand Abstr 2013:273–278. https://doi.org/10.1190/segam2013-0269.1
• 28. Zhou Y, Gao J, Chen W, Frossard P (2016) Seismic simultaneous source separation via patchwise sparse representation. IEEE Trans Geosci Remote Sens 54(9):5271–5284. https://doi.org/10.1109/TGRS.2016.2559514
• 29. Zu S et al (2017a) Iterative deblending of simultaneous-source data using a coherency-pass shaping operator. Geophys J Int 211(1):541–557. https://doi.org/10.1093/gji/ggx324
• 30. Zu S, Zhou H, Chen H, Zheng H, Chen Y (2017b) Two field trials for deblending of simultaneous source surveys: Why we failed and why we succeeded? J Appl Geophys 143:182–194. https://doi.org/10.1016/j.jappgeo.2017.06.002
• 31. Zu S, Cao J, Qu S, Chen Y (2019a) Iterative deblending for simultaneous source data using the deep neural network. Geophysics 85(2):V131–V141. https://doi.org/10.1190/geo2019-0319.1
• 32. Zu S, Zhou H, Wu R, Jiang M, Chen Y (2019b) Dictionary learning based on dip patch selection training for random noise attenuation. Geophysics 84(3):V169–V183. https://doi.org/10.1190/geo2018-0596.1

Uwagi

PL

Opracowanie rekordu ze środków MEiN, umowa nr SONP/SP/546092/2022 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2022-2023).