Cite this article
2
Download
50
Views
Journal Browser
Volume | Year
Issue
Search
News and Announcements
View All
ARTICLE

CNN-based adaptive subtraction for the removal of seismic multiples

ZHONGXIAO LI1,2 FEI XIE1,2 JIAHUI MA3 ZHEN QI3 YIBO WANG4
Show Less
1 State Key Laboratory of Shale Oil and Gas Enrichment Mechanisms and Effective Development, Beijing 100083, P.R. China,
2 Sinopec Key Laboratory of Seismic Elastic Wave Technology, Beijing 100083, P.R. China,
3 Department of Electronic Engineering, School of Electronic Information, Qingdao University, Qingdao 266071, Shandong Province, P.R. China,
4 Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100029, P.R. China,
JSE 2023, 32(2), 169–184;
Submitted: 16 November 2022 | Accepted: 10 March 2023 | Published: 1 April 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

Li, Z.X., Xie, F., Ma, J.H., Qi, Z. and Wang, Y., 2023. CNN-based adaptive subtraction for the removal of seismic multiples. Journal of Seismic Exploration, 32: 169-184. In seismic data processing primaries are usually distorted by multiples which need to be removed in advance before seismic imaging. After multiple modelling, adaptive subtraction is essential for removing multiples successfully and can be expressed as a problem of linear regression (LR) with L1 norm minimization constraint on primaries or support vector regression (SVR). Compared to the LR-based method, the SVR-based method achieves better separation of primaries and multiples since it transforms the modelled multiples nonlinearly for a better match with the true multiples in every 2D data window. However, the LR- or SVR-based method may harm primaries or cause residual multiples in complex subsurface media. In this paper a deep convolutional neural network (CNN) is constructed to better express the complicated mismatches between the modelled multiples (input data) and true multiples of the original data (label) than the LR or SVR model. To avoid overfitting to the original data and preserve primaries the L1 norm minimization constraint on primaries and L2 norm minimization constraint on CNN coefficients are used in the optimization problem. During CNN training multiple 2D data windows constructed with one or several gathers are used simultaneously to avoid overfitting. The trained CNN is used in the corresponding training data to remove multiples and then the same flowchart with CNN is used in other gathers. The proposed CNN-based method extracts high-level features of the modelled multiples to remove multiples. It is demonstrated in the synthetic and field data examples that the proposed CNN-based method can better remove multiples and preserve primaries than the LR- or SVR-based method.

Keywords
adaptive subtraction
seismic multiple removal
convolutional neural network
References
  1. Abma, R., Kabir, N., Matson, K.H., Michell, S., Shaw, S.A. and McLain, B., 2005.
  2. Comparisons of adaptive subtraction methods for multiple attenuation. The LeadingEdge, 24: 277-280.
  3. Berkhout, A.J. and Verschuur, D.J., 1997. Estimation of multiple scattering by iterativeinversion, Part I: Theoretical considerations. Geophysics, 62: 1586-1595.
  4. Bishop, K., Keliher, J., Paffenholz, J., Stoughton, D., Michell, S., Ergas, R. and Hadidi,
  5. M., 2001. Investigation of vendor demultiple technology for complex subsalt geology.
  6. Expanded Abstr, 71 st Ann. Internat. SEG Mtg., San Antonio: 1273-1276.
  7. Bugge, A.J., Evensen, A.K., Lie, J.E. and Nilsen, E.H., 2021. Demonstrating multipleattenuation with model-driven processing using neural networks. The Leading Edge,40: 831-836.
  8. Cheng, G., Zhou, P. and Han, J., 2016. Learning rotation-invariant convolutional neuralnetworks for object detection in VHR optical remote sensing images. IEEE Transact.Geosci. Remote Sens., 54: 7405-7415.
  9. Donno, D., 2011. Improving multiple removal using least-squares dip filters andindependent component analysis. Geophysics, 76(5): V91-V104.
  10. Gao, Z., Pan, Z., Zuo, C., Gao, J. and Xu, Z., 2019. An optimized deep networkrepresentation of multimutation differential evolution and its application in seismicinversion. IEEE Transact. Geosci. Remote Sens., 57: 4720-4734.
  11. Geng, Z., Wu, X., Shi, Y. and Fomel, S., 2020. Deep learning for relative geologic timeand seismic horizons. Geophysics, 85: 1-47.
  12. Guitton, A. and Verschuur, D.J., 2004. Adaptive subtraction of multiples using the L1-norm. Geophys.Prosp., 52: 27-38.
  13. Guitton, A., 2005. Multiple attenuation in complex geology with a pattern-basedapproach. Geophysics, 70(Issue No. ???):: V97-V 107.
  14. He, K., Zhang, X., Ren, S. and Sun, J., 2016. Deep residual learning for imagerecognition. IEEE Conf. Comput. Vis. Pattern Recognit., Las Vegas: 770-778.
  15. Jiao, Q., Ma, J., Chi, L., Liao, Z. and Li, C., 2021. Surface-related multiple attenuationbased on deep learning. 82nd EAGE Conf., Amsterdam: 1-5.
  16. Jin, K.H., McCann, M.T., Froustey, E. and Unser, M., 2017. Deep convolutional neuralnetwork for inverse problems in imaging. IEEE Transact. Image Process., 26: 4509-
  17. Kaplan, S.T. and Innanen, K.A., 2008. Adaptive separation of free-surface multiplesthrough independent component analysis. Geophysics, 73(4): V29-V36.
  18. Kaur, H., Pham, N. and Fomel, S., 2019. Seismic data interpolation using CycleGAN.
  19. Expanded Abstr., 89 th Ann. Internat. SEG Mtg., San Antonio: 2202-2206.
  20. Kingma, D.P. and Ba, J.L., 2015. Adam: A Method for stochastic optimization. Internat.Conf. Learning Representat., San Diego: 1-15.
  21. Kumar, A., Hampson, G. and Rayment, T., 2021. Adaptive subtraction using aconvolutional neural network. First Break, 39: 35-45.
  22. LeCun, Y., Bengio, Y. and Hinton, G., 2015. Deep learning. Nature, 521: 436-444.
  23. Liu, D., Wang, W., Wang, X., Wang, C., Pei, J., and Chen, W., 2020. Poststack seismicdata denoising based on 3-D convolutional neural network. IEEE Transact. Geosci.Remote Sens., 58: 1598-1629.
  24. Liu, J., Lu, W., and Zhang, Y., 2017. Adaptive multiple subtraction based on sparsecoding. IEEE Transactions on Geoscience and Remote Sensing, 55: 1318-1324.
  25. Li, Z.-X., 2020. Adaptive multiple subtraction based on support vector regression.Geophysics, 85(3): V57-V69.
  26. Li, Z. and Gao, H., 2020. Feature extraction based on the convolutional neural networkfor adaptive multiple subtraction. Marine Geophys. Res., 41: 10.
  27. Li, Z.-X. and Li, Z.-C., 2018. Accelerated 3D blind separation of convolved mixturesbased on the fast iterative shrinkage thresholding algorithm for adaptive multiplesubtraction. Geophysics, 83(1): V99-V113.
  28. Li, Z.-X. and Li, Z.-C., 2017. Accelerated parabolic Radon domain 2D adaptive multiplesubtraction with fast iterative shrinkage thresholding algorithm and its application inparabolic Radon domain hybrid demultiple method. J. Appl. Geophys., 143: 86-102.
  29. Li, Z., Sun, N., Gao, H., Qin, N. and Li, Z., 2021. Adaptive subtraction based on U-netfor removing seismic multiples. IEEE Transact. Geosci. Remote Sens., 59: 9796-9812.
  30. Mousa, W.A., 2014. Imaging of the SEG/EAGE salt model seismic data using sparse f-xfinite-impulse-response wavefield extrapolation filters. IEEE Transact. Geosci.Remote Sens., 52: 2700-2714.
  31. Neelamani, R.N., Baumstein, A. and Ross, W.S., 2010. Adaptive subtraction usingcomplex-valued curvelet transforms. Geophysics, 75(2): V51-V60.
  32. Schmidhuber, J., 2015. Deep learning in neural networks: An overview. Neural Netw.,61: 85-117.
  33. Siahkoohi, A., Verschuur, D.J. and Herrmann, F.J., 2019. Surface-related multipleelimination with deep learning. Expanded Abstr., 89th Ann. Internat. SEG Mtg., SanAntonio: 4629-4634.
  34. Spitz, S., 1999. Pattern recognition, spatial predictability, and subtraction of multipleevents. Geophysics, 18: 55-58.
  35. Verschuur, D.J. and Berkhout, A.J., 1997. Estimation of multiple scattering by iterativeinversion; Part II, Practical aspects and examples. Geophysics, 62: 1596-1611.
  36. Verschuur, D.J., 2006. Seismic Multiple Removal Techniques: Past, Present and Future.EAGE, Houten.
  37. Yang, F. and Ma, J., 2019. Deep-learning inversion: A next-generation seismic velocitymodel building method. Geophysics, 84(4): R585-R584.
  38. Yuan, S., Liu, J., Wang, S., Wang, T. and Shi, P., 2018. Seismic waveform classificationand first-break picking using convolution neural networks. IEEE Geosci. RemoteSens. Lett., 15: 272-276.
  39. Yu, S., Ma, J. and Wang, W., 2019. Deep learning for denoising. Geophysics, 84(6):V333-V350.
  40. Zhang, D., Leeuw, M. and Verschuur, E., 2021. Deep learning-based seismic surface-related multiple adaptive subtraction with synthetic primary labels. Expanded Abstr.,
  41. First Internat. Mtg. Appl. Geosci. Energy, Denver: 2844-2848.
  42. 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.
Share
Back to top
Journal of Seismic Exploration, Electronic ISSN: 0963-0651 Print ISSN: 0963-0651, Published by AccScience Publishing