Geophysics and Geophysical Exploration (지구물리와물리탐사)

Korean Society of Earth and Exploration Geophysicists (한국지구물리·물리탐사학회)
권호 표지
DOI QR코드

DOI QR Code

Trace-based Interpolation Using Machine Learning for Irregularly Missing Seismic Data

불규칙한 빠짐을 포함한 탄성파 탐사 자료의 머신러닝을 이용한 트레이스 기반 내삽

  • Zeu Yeeh (Department of Earth Resources and Environmental Engineering, Hanyang University) ;
  • Jiho Park (Department of Earth Resources and Environmental Engineering, Hanyang University) ;
  • Soon Jee Seol (Department of Earth Resources and Environmental Engineering, Hanyang University) ;
  • Daeung Yoon (Department of Energy and Resources Engineering, Chonnam National University) ;
  • Joongmoo Byun (Department of Earth Resources and Environmental Engineering, Hanyang University)
  • 이재우 (한양대학교 자원환경공학과) ;
  • 박지호 (한양대학교 자원환경공학과) ;
  • 설순지 (한양대학교 자원환경공학과) ;
  • 윤대웅 (전남대학교 에너지자원공학과) ;
  • 변중무 (한양대학교 자원환경공학과)
  • Received : 2023.04.12
  • Accepted : 2023.05.01
  • Published : 2023.05.31
Download PDF
⟨ Previous Next ⟩

Abstract

Recently, machine learning (ML) techniques have been actively applied for seismic trace interpolation. However, because most research is based on training-inference strategies that treat missing trace gather data as a 2D image with a blank area, a sufficient number of fully sampled data are required for training. This study proposes trace interpolation using ML, which uses only irregularly sampled field data, both in training and inference, by modifying the training-inference strategies of trace-based interpolation techniques. In this study, we describe a method for constructing networks that vary depending on the maximum number of consecutive gaps in seismic field data and the training method. To verify the applicability of the proposed method to field data, we applied our method to time-migrated seismic data acquired from the Vincent oilfield in the Exmouth Sub-basin area of Western Australia and compared the results with those of the conventional trace interpolation method. Both methods showed high interpolation performance, as confirmed by quantitative indicators, and the interpolation performance was uniformly good at all frequencies.

최근에 활발히 적용되고 있는 머신러닝 기반 탄성파 내삽 기법들은 대부분 모음 자료를 2차원 영상화 하여 빠짐을 채우는 방법으로 하는 훈련(training)-추론(inference) 전략에 기초하므로 완벽히 채워진 다수의 모음자료가 훈련을 위해 필요하게 된다. 이 연구는 이와는 달리 트레이스 기반 내삽을 수행하는 내삽 기술의 훈련-추론 전략을 기본으로, 불규칙한 빠짐이 있는 현장자료 만을 이용하여 훈련-추론을 모두 수행할 수 있는 머신러닝을 이용한 트레이스 기반 불규칙한 빠짐의 내삽 기술을 제시하였다. 이 연구에서는 불규칙한 빠짐이 있는 자료를 훈련과 추론에 체계적으로 사용하는 최대 연속빠짐 간격에 따라 정해지는 네트워크를 구성하는 방법 및 훈련하는 방법을 기술하였다. 또한, 서호주 Exmouth Sub-basin 지역의 Vincent 유전에서 얻어진 시간 참반사 보정된 탄성파 자료에 개발된 방법을 적용한 후, 예측 결과를 전통적인 내삽 방법의 결과와 비교 및 분석하였다. 신호대잡음비나 구조유사성과 같은 정량적인 지표를 통해 두 방법 모두 내삽 성능이 높은 것을 확인하였으며, 모든 주파수 대역에서도 골고루 좋은 결과를 보임을 확인하였다.

Keywords

Acknowledgement

이 논문(또는 특허 등)은(는) 2021년도 정부(과학기술정보통신부)의 재원으로 한국연구재단의 지원을 받아 수행된 연구임(No. 2021R1A2C2014315).

References

  1. Abma, R., and Kabir, N., 2006, 3D interpolation of irregular data with a POCS algorithm, GEOPHYSICS, 71(6), E91-E97, doi: https://doi.org/10.1190/1.2356088
  2. Bourlard, H., and Kamp, Y., 1988, Auto-association by multilayer perceptrons and singular value decomposition, Biol. Cybern., 59(4-5), 291-294, doi: https://doi.org/10.1007/BF00332918
  3. Choi, J., Byun, J., Seol, S. J., and Kim, Y., 2016, Wavelet-based multicomponent matching pursuit trace interpolation, Geophys. J. Int., 206(3), 1831-1846, doi: https://doi.org/10.1093/gji/ggw246
  4. Choi, J, Song, Y., Choi, J., Byun, J., Seol, S. J., Kim, K., and Lee, J., 2017, Trace Interpolation using Model-constrained Minimum Weighted Norm Interpolation, Geophys. Geophys. Explor., 20(2), 78-87, doi: https://doi.org/10.7582/GGE.2017.20.2.078 (In Korean with Enghlish abstract)
  5. Dong, C., Loy, C. C., He, K., and Tang, X., 2016, Image Super-Resolution Using Deep Convolutional Networks, IEEE Trans. Pattern Anal. Mach. Intell., 38(2), 295-307, doi: https://doi.org/10.1109/TPAMI.2015.2439281
  6. Dou, Y., Li, K., Duan, H., Li, T., Dong, L., and Huang, Z., 2022, MDA GAN: Adversarial-Learning-based 3-D Seismic Data Interpolation and Reconstruction for Complex Missing, IEEE Trans. Geosci. Remote Sens., 61, 1-14, doi: https://doi.org/10.1109/TGRS.2023.3249476
  7. Fang, W., Fu, L., Wu, M., Yue, J., and Li, H., 2023, Irregularly sampled seismic data interpolation with self-supervised learning, GEOPHYSICS, 88(3), 1-43, doi: https://doi.org/10.1190/geo2022-0586.1
  8. Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., WardeFarley, D., Ozair, S., Courville, A., and Bengio, Y., 2020, Generative adversarial networks, Commun. ACM, 63(11), 139-144, doi: https://doi.org/10.1145/3422622
  9. Hore, A., and Ziou, D., 2010, Image Quality Metrics: PSNR vs. SSIM, 2010 20th International Conference on Pattern Recognition, IEEE, 2366-2369, doi: https://doi.org/10.1109/ICPR.2010.579
  10. Jia, Y., and Ma, J., 2017, What can machine learning do for seismic data processing? An interpolation application, GEOPHYSICS, 82(3), V163-V177, doi: https://doi.org/10.1190/geo2016-0300.1
  11. Kim, B., Jeong, S., and Byun, J., 2015, Trace interpolation for irregularly sampled seismic data using curvelet-transform-based projection onto convex sets algorithm in the frequency-wavenumber domain, J. Appl. Geophys., 118, 1-14, doi: https://doi.org/10.1016/j.jappgeo.2015.04.007
  12. Lilly, J. M., and Olhede, S. C., 2012, Generalized Morse Wavelets as a Superfamily of Analytic Wavelets, IEEE Trans. Signal Process., 60(11), 6036-6041, doi: https://doi.org/10.1109/TSP.2012.2210890
  13. Liu, B., and Sacchi, M. D., 2004, Minimum weighted norm interpolation of seismic records, GEOPHYSICS, 69(6), 1560-1568, doi: https://doi.org/10.1190/1.1836829
  14. Oliveira, D. A. B., Ferreira, R. S., Silva, R., and Brazil, E. V., 2018, Interpolating Seismic Data With Conditional Generative Adversarial Networks, IEEE Geosci. Remote Sens. Lett., 15(12), 1952-1956, doi: https://doi.org/10.1109/LGRS.2018.2866199
  15. Pan, S., Chen, K., Chen, J., Qin, Z., Cui, Q., and Li, J., 2020, A partial convolution-based deep-learning network for seismic data regularization1, Comput. Geosci., 145, 104609, doi: https://doi.org/10.1016/J.CAGEO.2020.104609
  16. Park, J., Yeeh, Z., Seol, S. J., and Byun, J., 2022, Seismic Data Interpolation Using Attention-Based Deep Learning, 83rd EAGE Annual Conference & Exhibition, European Association of Geoscientists & Engineers, 2, 1-5, doi: https://doi.org/10.3997/2214-4609.202210245
  17. Park, J., Choi, J., Seol, S. J., Byun, J., and Kim, Y., 2021, A method for adequate selection of training data sets to reconstruct seismic data using a convolutional U-Net, GEOPHYSICS, 86(5), V375-V388, doi: https://doi.org/10.1190/geo2019-0708.1
  18. Spitz, S., 1991, Seismic trace interpolation in the F-X domain, GEOPHYSICS, 56(6), 785-794, doi: https://doi.org/10.1190/1.1443096
  19. Trad, D., 2009, Five-dimensional interpolation: Recovering from acquisition constraints, GEOPHYSICS, 74(6), V123-V132, doi: https://doi.org/10.1190/1.3245216
  20. Trad, D., 2014, Five-dimensional interpolation: New directions and challenges, CSEG Rec., 39(3), 22-29, https://csegrecorder.com/articles/view/five-dimensional-interpolation-newdirections-and-challenges
  21. Vassallo, M., Ozbek, A., Ozdemir, K., and Eggenberger, K., 2010, Crossline wavefield reconstruction from multicomponent streamer data: Part 1 - Multichannel interpolation by matching pursuit (MIMAP) using pressure and its crossline gradient, GEOPHYSICS, 75(6), WB53-WB67, doi: https://doi.org/10.1190/1.3496958
  22. Wang, B., Zhang, N., Lu, W., Geng, J., and Huang, X., 2020a, Intelligent Missing Shots' Reconstruction Using the Spatial Reciprocity of Green's Function Based on Deep Learning, IEEE Trans. Geosci. Remote Sens., 58(3), 1587-1597, doi: https://doi.org/10.1109/TGRS.2019.2947085
  23. Wang, Y., Wang, B., Tu, N., and Geng, J., 2020b, Seismic trace interpolation for irregularly spatial sampled data using convolutional autoencoder, GEOPHYSICS, 85(2), V119-V130, doi: https://doi.org/10.1190/geo2018-0699.1
  24. Wang, Z., Bovik, A. C., Sheikh, H. R., and Simoncelli, E. P., 2004, Image quality assessment: from error visibility to structural similarity, IEEE Trans. Image Process., 13(4), 600-12, doi: https://doi.org/10.1109/tip.2003.819861
  25. Xu, S., Zhang, Y., and Lambare, G., 2010, Antileakage Fourier transform for seismic data regularization in higher dimensions, GEOPHYSICS, 75(6), WB113-WB120, doi: https://doi.org/10.1190/1.3507248
  26. Yeeh, Z., Byun, J., and Yoon, D., 2020a, Crossline interpolation with the traces-to-trace approach using machine learning, SEG Technical Program Expanded Abstracts 2020, Society of Exploration Geophysicists, 2020-Octob, 1656-1660, doi: https://doi.org/10.1190/segam2020-3428348.1
  27. Yeeh, Z., Song, Y., Byun, J., Seol, S. J., and Kim, K. Y., 2020b, Regularization of multidimensional sparse seismic data using Delaunay tessellation, J. Appl. Geophys., 174, 103877, doi: https://doi.org/10.1016/J.JAPPGEO.2019.103877
  28. Yoon, D., Yeeh, Z., and Byun, J., 2021, Seismic Data Reconstruction Using Deep Bidirectional Long Short-Term Memory with Skip Connections, IEEE Geosci. Remote Sens. Lett., 18(7), 1298-1302, doi: https://doi.org/10.1109/LGRS.2020.2993847