Geophysics and Geophysical Exploration (지구물리와물리탐사)

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

DOI QR Code

Synthetic Training Data Generation for Fault Detection Based on Deep Learning

딥러닝 기반 탄성파 단층 해석을 위한 합성 학습 자료 생성

  • Choi, Woochang (Department of Energy Resources Engineering, Inha University) ;
  • Pyun, Sukjoon (Department of Energy Resources Engineering, Inha University)
  • 최우창 (인하대학교 에너지자원공학과) ;
  • 편석준 (인하대학교 에너지자원공학과)
  • Received : 2021.07.30
  • Accepted : 2021.08.25
  • Published : 2021.08.31
Download PDF
⟨ Previous Next ⟩

Abstract

Fault detection in seismic data is well suited to the application of machine learning algorithms. Accordingly, various machine learning techniques are being developed. In recent studies, machine learning models, which utilize synthetic data, are the particular focus when training with deep learning. The use of synthetic training data has many advantages; Securing massive data for training becomes easy and generating exact fault labels is possible with the help of synthetic training data. To interpret real data with the model trained by synthetic data, the synthetic data used for training should be geologically realistic. In this study, we introduce a method to generate realistic synthetic seismic data. Initially, reflectivity models are generated to include realistic fault structures, and then, a one-way wave equation is applied to efficiently generate seismic stack sections. Next, a migration algorithm is used to remove diffraction artifacts and random noise is added to mimic actual field data. A convolutional neural network model based on the U-Net structure is used to verify the generated synthetic data set. From the results of the experiment, we confirm that realistic synthetic data effectively creates a deep learning model that can be applied to field data.

탄성파 자료에서의 단층 해석은 기계학습을 적용하기 매우 적합한 분야라고 할 수 있다. 결과적으로 다양한 형태의 기계학습 기반 단층 해석 기술들이 개발되고 있으며, 특히 합성 자료를 사용해 기계학습 모델을 훈련시키는 연구들이 중점적으로 수행되고 있다. 합성 자료를 사용할 경우 기계학습 모델을 훈련시키기 위한 대량의 자료를 확보하기가 용이하고, 정확한 단층 구조 라벨을 함께 제작할 수 있다는 장점이 있다. 합성 자료로 훈련시킨 모델을 사용해 현장 자료를 해석하기 위해서는 모델 훈련에 사용한 합성 자료가 지질학적으로 현실적이어야 한다. 이 연구에서는 실제 현장 자료와 유사한 합성 자료 제작을 위한 기술을 소개한다. 먼저 현실적인 단층 구조가 포함된 반사계수 모델을 제작한 후 일방향 파동 방정식 모델링을 적용해 효율적으로 겹쌓기 단면을 생성한다. 생성된 겹쌓기 단면에 참반사보정을 적용해 회절파의 영향을 제거하고, 무작위 잡음을 추가함으로써 현장 자료와 비슷한 형태의 합성 자료를 생성할 수 있다. 생성한 합성 자료를 U-Net 구조의 합성곱 신경망 모델에 적용하여 검증한 결과, 현실적으로 만들어진 합성 자료는 현장 자료에 적용이 가능한 딥러닝 모델을 효과적으로 훈련시킬 수 있다는 것을 확인하였다.

Keywords

Acknowledgement

이 연구는 한국해양과학기술원 주요사업(PE99941)의 지원으로 수행되었습니다.

References

  1. An, Y., Guo, J., Ye, Q., Childs, C., Walsh, J., and Dong, R., 2021, Deep convolutional neural network for automatic fault recognition from 3D seismic datasets, Comput. Geosci., 153, 104776, doi: 10.1016/j.cageo.2021.104776.
  2. Chang, D., Yong, X., Yang, W., Wang, Y., and Guo, T. C., 2019, U_Net & Residual Neural Networks for Seismic Fault Interpretation, 81st Ann. Internat. Mtg., Eur. Assn. Geosci. Eng., Conference Proceedings, doi: 10.3997/2214-4609.201901387.
  3. Cunha, A., Pochet, A., Lopes, H., and Gattass, M., 2020, Seismic fault detection in real data using transfer learning from a convolutional neural network pre-trained with synthetic seismic data, Comput. Geosci., 135, 104344, doi: 10.1016/j.cageo.2019.104344.
  4. Gazdag, J., 1978, Wave equation migration with the phase-shift method, Geophysics, 43(7), 1342-1351, doi: 10.1190/1.1440899.
  5. Guillon, S., Joncour, F., Goutorbe, P., and Castanie, L., 2019, Reducing training dataset bias for automatic fault detection, 89th Ann. Internat. Mtg., Soc. Expl. Geophys., Expanded Abstract, 2423-2427, doi: 10.1190/segam2019-3216557.1.
  6. Hale, D., 2014, Seismic image processing for geologic faults, https://github.com/dhale/ipf (July 29, 2021, Accessed)
  7. Kallweit, R. S., and Wood, L. C., 1982, The limits of resolution of zero-phase wavelets, Geophysics, 47(7), 1035-1046, doi: 10.1190/1.1441367.
  8. Lee, M. W., and Suh, S. Y., 1985, Optimization of one-way wave equations, Geophysics, 50(10), 1634-1637, doi: 10.1190/1.1441853.
  9. Li, S., Yang, C., Sun, H., and Zhang, H., 2019, Seismic fault detection using an encoder-decoder convolutional neural network with a small training set, J. Geophys. Eng., 16(1), 175-189, doi: 10.1093/jge/gxy015.
  10. Liu, N., He, T., Tian, Y., Wu, B., Gao, J., and Xu, Z., 2020, Common-azimuth seismic data fault analysis using residual UNet, Interpretation, 8(3), SM25-SM37, doi: 10.1190/INT2019-0173.1.
  11. Pochet, A., Diniz, P. H., Lopes, H., and Gattass, M., 2019, Seismic fault detection using convolutional neural networks trained on synthetic poststacked amplitude maps, IEEE Geosci. Remote Sens. Lett., 16(3), 352-356, doi: 10.1109/LGRS.2018.2875836.
  12. Qi, J., Lyu, B., Wu, X., and Marfurt, K., 2020, Comparing convolutional neural networking and image processing seismic fault detection methods, 90th Ann. Internat. Mtg., Soc. Expl. Geophys., Expanded Abstract, 1111-1115, doi: 10.1190/segam2020-3428171.1.
  13. Qi, J., Machado, G., and Marfurt, K., 2017, A workflow to skeletonize faults and stratigraphic features, Geophysics, 82(4), O57-O70, doi: 10.1190/geo2016-0641.1.
  14. Ren, Y., Nie, L., Yang, S., Jiang, P., and Chen, Y., 2021, Building complex seismic velocity models for deep learning inversion, IEEE Access, 9, 63767-63778, doi: 10.1109/ACCESS.2021.3051159.
  15. Ricker, N., 1953, Wavelet contraction, wavelet expansion and the control of seismic resolution, Geophysics, 18, 769-792, doi: 10.1190/1.1437927.
  16. Ronneberger, O., Fischer, P., and Brox, T., 2015, U-net: Convolutional networks for biomedical image segmentation, Med Image Comput Comput Assist Interv, 234-241, doi: 10.1007/978-3-319-24574-4_28.
  17. Wu, X., Geng, Z., Shi, Y., Pham, N., Fomel, S., and Caumon, G., 2020, Building realistic structure models to train convolutional neural networks for seismic structural interpretation, Geophysics, 85(4), WA27-WA39, doi: 10.1190/geo2019-0375.1.
  18. Wu, X., Liang, L., Shi, Y., and Fomel, S., 2019, FaultSeg3D: Using synthetic data sets to train an end-to-end convolutional neural network for 3D seismic fault segmentation, Geophysics, 84(3), IM35-IM45, doi: 10.1190/geo2018-0646.1.
  19. Xiong, W., Ji, X., Ma, Y., Wang, Y., AlBinHassan, N. M., Ali, M. N., and Luo, Y., 2018, Seismic fault detection with convolutional neural network, Geophysics, 83(5), O97-O103, doi: 10.1190/geo2017-0666.1.
  20. Yilmaz, O., 2001, Seismic data analysis, Tulsa, OK: Society of exploration geophysicists, doi: 10.1190/1.9781560801580.
  21. Yuan, C., Cai, M., Lu, F., Li, H., and Li, G., 2020, 3D Fault Detection Based on GCS-Net, 82nd Ann. Internat. Mtg., Eur. Assn. Geosci. Eng., Conference Proceedings, doi: 10.3997/2214-4609.202010699.