ARTICLE

Seismic random noise suppression using denoising autoencoder

HUI SONG MENGHUA FANG CHENG ZHOU HOUQIANG GAO
Show Less
Sinopec Geophysical Research Institute, Jiangning District, Nanjing 211103, P.R. China,
JSE 2022, 31(3), 203–218;
Submitted: 15 August 2021 | Accepted: 4 January 2022 | Published: 1 June 2022
© 2022 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

Song, H., Fang, M.H., Zhou, C. and Gao, H.Q., 2022. Seismic random noise suppression using denoising autoencoder. Journal of Seismic Exploration, 31: 203-218. In the seismic data acquisition phase, random noise will inevitably be introduced as undesirable components, which is not conducive to subsequent seismic signal processing and imaging tasks. In this article, we propose a denoising framework based on denoising autoencoder (DAE) for seismic random noise suppression. DAE can directly reconstruct noise-free seismic data from noisy seismic data in an unsupervised learning manner. The entire seismic data reconstruction requires three phases: corrupting phase, encoding phase, and decoding phase. In the corrupting phase, the original input is randomly corrupted, which helps the designed network to capture robust features. In the encoding phase, the corrupted input is encoded as a compressed representation that contains the important content of the seismic data. In the decoding stage, the compressed representation is decoded into reconstructed data. We choose the mean square error as the loss function, which is minimized by the back propagation algorithm to update the network parameters. The application of synthetic and real seismic data proves the effectiveness of the proposed method in suppressing seismic random noise.

Keywords
seismic data
random noise
unsupervised learning
denoising autoencoder
References
  1. Abma, R. and Claerbout, J., 1995. Lateral prediction for noise attenuation by t-x and f-xtechniques. Geophysics, 60(6): 1887-1896.
  2. Chen, Y. and Ma, J., 2014. Random noise attenuation by f-x empirical modedecomposition predictive filtering. Geophysics, 79(3): V81-V91.
  3. Chen, Y., Ma, J. and Fomel, S., 2016a. Double sparsity dictionary for seismic noiseattenuation. Geophysics, 81(2): V103-V116.
  4. Chen, Y., Zhang, D., Jin, Z., Chen, X., Zu, S., Huang, W. and Gan, S., 2016b.
  5. Simultaneous denoising and reconstruction of 5-D seismic data via damped rank-reduction method. Geophys. J. Internat., 206(3): 1695-1717.
  6. Chen, Y., Zhang, M., Bai, M. and Chen, W., 2019. Improving the signal-to-noise ratio ofseismological datasets by unsupervised machine learning. Seismol. Res. Lett., 90(4):1552-1564.
  7. Deighan, A. and Watts, D., 1997. Ground-roll suppression using the wavelet transform.Geophysics, 62(6): 1896-1903.
  8. Gao, Z., Pan, Z. and Gao, J., 2016. Multimutation differential evolution algorithm and itsapplication to seismic inversion. IEEE Transact. Geosci. Remote Sens., 54(6): 3626-
  9. Herrmann, F., Boniger, U. and Verschuur, D., 2007. Non-linear primary-multipleseparation with directional curvelet frames. Geophys. J. Internat., 170(2): 781-799.
  10. Herrmann, F. and Hennenfent, G., 2008. Non-parametric seismic data recovery withcurvelet frames. Geophys. J. Internat., 173(1): 233-248.
  11. Kazemi, N., Bongajum, E. and Sacchi, M., 2016. Surface-consistent sparse multichannelblind deconvolution of seismic signals. IEEE Transact. Geosci. Remote Sens., 54(6):3200-3207.
  12. Kingma, D.P. and Ba, J., 2015. Adam: A method for stochastic optimization. CoRR,abs/1412.6980.
  13. Lin, H., Li, Y., Yang, B. and Ma, H., 2013. Random denoising and signal nonlinearityapproach by time-frequency peak filtering using weighted frequency reassignment.Geophysics, 78(6): V229-V237.
  14. Liu, W., Cao, S., Jin, Z., Wang, Z. and Chen, Y., 2018. A novel hydrocarbon detectionapproach via high-resolution frequency-dependent AVO inversion based onvariational mode decomposition. IEEE Transact. Geosci. Remote Sens., 56(4): 2007-
  15. Oropeza, V. and Sacchi, M., 2011. Simultaneous seismic data denoising andreconstruction via multichannel singular spectrum analysis. Geophysics, 76(3): V25-V32.
  16. Siahsar, M.A.N., Gholtashi, S., Kahoo, A.R., Chen, W. and Chen, Y., 2017. Data-drivenmultitask sparse dictionary learning for noise attenuation of 3D seismic data.Geophysics, 82(6): V385-V396.
  17. Song, H., Chen, W., Zhang, H., Wang, Y. and Xue, Y., 2020. Sandstone porosityprediction based on gated recurrent units. J. Seismic Explor., 29(4): 371-388.
  18. Wu, X., Liang, L., Shi, Y. and Fomel, S., 2019. Faultseg3D: Using synthetic data sets totrain an end-to-end convolutional neural network for 3D seismic fault segmentation.Geophysics, 84(3): IM35-IM45.
  19. Yu, S., Ma, J. and Wang, W., 2019. Deep learning for denoising. Geophysics, 84(6):V333-V350.
  20. 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(2): 272-276.
  21. Zhang, M., Liu, Y., Bai, M. and Chen, Y., 2019. Seismic noise attenuation usingunsupervised sparse feature learning. IEEE Transact. Geosci. Remote Sens., 57(12):9709-9723.
Share
Back to top
Journal of Seismic Exploration, Electronic ISSN: 0963-0651 Print ISSN: 0963-0651, Published by AccScience Publishing