AccScience Publishing / JSE / Article
Submit to JSE
Cite this article
1
Download
85
Views
Journal Browser
Volume | Year
Issue
Search
Forthcoming Issue
Current Issue
View All
News and Announcements
View All
ARTICLE

Efficient seismic numerical modelling technique using the YOLOv2-based expanding domain method

DAWOON LEE1 ,  SEUNGPYO CHOI2 ,  JONGHYUN LEE3 ,  WOOKEEN CHUNG4
Show Less
4 Department of Ocean Energy and Resources Engineering, Korea Maritime and Ocean University, Busan, South Korea.,
2 Department of Convergence Study on the Ocean Science and Technology, Ocean Science and Technology (OST) School, Korea Maritime and Ocean University, Busan, South Korea,
3 Department of Civil and Environmental Engineering and Water Resources Research Center, University of Hawai’i at Manoa, Honolulu, HI 96822, U.S.A.,
JSE 2022, 31(5), 425–449;
Submitted: 11 April 2022 | Accepted: 5 August 2022 | Published: 1 October 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/ )
Download PDF
Cite
XML
Abstract

D.W. Lee, S.P. Choi, J.H. Lee and W.K. Chung. Efficient seismic numerical modeling technique using the YOLOv2-based expanding domain method. Journal of Seismic Exploration, 31: 425-449. Wave equation-based seismic modeling has the advantage of simulating the exact full- wave propagation. However, it requires a great amount of computational resources, which becomes prohibitive when both the modeling domain size and the number of the time samples increase. Therefore, much research has been performed to enhance the computational efficiency of seismic numerical modeling. The expanding domain method is such one technique that improves the computational efficiency by identifying the domain extent where the wave propagation has not reached and excluding these domains from the calculation. In this work, we propose a new deep-learning based method that guide the determination of the computational domain. In order to establish the computational domain where the wave propagates from the snapshots. the deep learning-based obiect detection was emploved. The deep learning obiect detector used has two main components. The first one is a structure for the feature extraction layers based on ResNet-50. The second one is a structure for the detection of the wave propagation domain based on the You Only Look Once method, version 2 (YOLOv2). After the training, validation and test for the YOLOv2 obiect detector, the computational efficiency of our proposed method was compared with that of the widely used amplitude comparison-based expanding domain method. It was demonstrated that the computational efficiency of the YOLOv2 method was better when the number of modeling grids was large, and the efficiency in the largest number of grids was about 25.1 %.

Keywords
seismic numerical modeling
computational efficiencv
expanding domain method
deep learning object detection
YOLOv2
References
  1. Abdelkhalek, R., Calandra, H., Coulaud, O., Roman, J. and Latu, G., 2009. Fast seismicmodeling and reverse time migration on a GPU cluster. 2009 Internat. Conf. HighPerformance Comput. Simulat.: 36-43.
  2. Alkhalifah, T., 2000. An acoustic wave equation for anisotropic media, Geophysics, 65:1239-1250.
  3. Aminzadeh, F., Burkhard, N., Nicoletis, L., Rocca, F. and Wyatt, K., 1994. SEG/EAEG3-D modeling project, 2nd update. The Leading Edge, 13: 949-952.
  4. Barthakur, M. and Sarma, K.K., 2019. Semantic segmentation using K-means clusteringand deep learning in satellite image. 22nd Internat. Conf. Innovat. Electronics,Signal Process. Communicat. (IESC): 192-196.
  5. Carcione, J.M., Herman, G.C. and Ten Kroode, A.P.E., 2002. Seismic modelling,Geophysics, 67: 1304-1325.
  6. Carcione, J.M., Kosloff, D. and Kosloff, R., 1988. Viscoacoustic wave propagationsimulation in the earth, Geophysics, 53: 769-777.
  7. Cerjan, C., Kosloff, D., Kosloff, R. and Reshef, M., 1985. A nonreflecting boundarycondition for discrete acoustic and elastic wave equations, Geophysics, 50: 705-708.
  8. Chassagnon, G., Vakalopolou, M., Paragios, N. and Revel, M.P., 2020. Deep learning:definition and perspectives for thoracic imaging. Eur. Radiol., 30: 2021-2030.
  9. Deng, J., Dong, W., Socher, R., Li, L.J., Li, K. and Fei-Fei, L., 2009. Imagenet: A large-scale hierarchical image database. 2009 IEEE Conf. Comput. Vision Patt. Recognit.,248-255.
  10. Drossaert, F.H. and Giannopoulos, A., 2007. A nonsplit complex frequency-shifted PMLbased on recursive integration for FDTD modeling of elastic waves. Geophysics,72(2): T9-T17.
  11. Etgen, J.T. and O’Brien, M.J., 2007. Computational methods for large-scale 3D acousticfinite-difference modeling: A tutorial. Geophysics, 72(5): SM223-SM230.
  12. Fowler, P.J. and King, R., 2011. Modeling and reverse time migration of orthorhombicpseudo-acoustic P-waves. Expanded Abstr., 81st. Ann. Internat. SEG Mtg., SanAntonio: 190-195.
  13. Geiger, A., Lenz, P., Stiller, C. and Urtasun, R., 2013. Vision meets robotics: The kittidataset. Internat. J. Rob. Res., 32: 1231-1237.
  14. Girshick, R., 2015. Fast R-CNN. Proc. IEEE Internat. Conf. Comput. Vis.: 1440-1448.
  15. Girshick, R., Donahue, J., Darrell, T. and Malik, J., 2014. Rich feature hierarchies foraccurate object detection and semantic segmentation. Proc. IEEE Conf. Comput.Vis. Patt. Recognit.. 580-587.
  16. Graves, R.W., 1996. Simulating seismic wave propagation in 3D elastic media usingstaggered-grid finite differences. Bull. Seismol. Soc. Am., 86: 1091-1106.
  17. Ha, W. and Shin, C., 2012. Efficient Laplace-domain modeling and inversion using anaxis transformation technique. Geophysics, 77(4): R141-R148.
  18. He, K., Gkioxari, G., Dollar, P. and Girshick, R., 2017. Mask R-CNN. Proc. IEEEInternat. Conf. Comput. Vis. (ICCV): 2961-2969.
  19. He, K., Zhang, X., Ren, S. and Sun, J., 2016. Deep residual learning for imagerecognition. Proc. IEEE Conf. Comput. Vis. Patt. Recognit.: 770-778.
  20. Itakura, K. and Hosoi, F., 2020. Automatic tree detection from three-dimensional imagesreconstructed from 360 spherical camera using YOLOv2. Remote Sens., 12: 988.
  21. Jun, H., Jou, H.T., Kim, C.H., Lee, S.H. and Kim, H.J., 2020. Random noise attenuationof sparker seismic oceanography data with machine learning, Ocean Sci., 16: 1367-
  22. Kim, S., Seol, S.J., Byun, J., Park, J. and Oh, S., 2020. Extraction of diffraction eventsfrom seismic data using deep learning-based approach. Expanded Abstr., 90th Ann.Internat. SEG Mtg., Houston: 2840-2844.
  23. Kingma, D.P. and Ba, J., 2014. Adam: A method for stochastic optimization. arXivpreprint arXiv: 1412.6980.
  24. Komatitsch, D. and Tromp, J., 1999. Introduction to the spectral element method forthree-dimensional seismic wave propagation, Geophys. J. Internat., 139: 806-822.
  25. Levander, A.R., 1988. Fourth-order finite-difference P-SV seismograms, Geophysics,53: 1425-1436.
  26. 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.Engineer., 16: 175-189.
  27. Lin, T.Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollar, P. and
  28. Zitnick, C.L., 2014. Microsoft COCO: Common objects in context. In: Fleet, D.,
  29. Pajdla, T., Schiele, B. and Tuytelaars ,T. (Eds.), European Conf. Comput. Vis.,
  30. Lecture Notes in Computer Science, Vol. 8693. Springer Verlag, Cham.: 740-755.
  31. Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C.Y. and Berg, A.C., 2016.
  32. SSD: Single shot multibox detector. In: Leibe, B., Matas, J., Sebe, N., Welling, M.(Eds.), European Conf. Comput, Vis.. Springer Verlag, Cham.: 21-37.
  33. Loey, M., Manogaran, G., Taha, M.H.N. and Khalifa, N.E.M., 2021. Fighting against
  34. COVID-19: A novel deep learning model based on YOLO-v2 with ResNet-50 formedical face mask detection. Sustain. Cities Soc., 65: 102600.
  35. Martin, G.S., Wiley, R. and Marfurt, K.J., 2006. Marmousi-2: An elastic upgrade for___Marmousi. The Leading. Edge, 25: 156-166.
  36. Ozdenvar, T. and McMechan, G.A., 1997. Algorithms for staggered-grid computationsfor poroelastic, elastic, acoustic, and scalar wave equations, Geophys. Prosp., 45:403-420.
  37. Padilla, R., Netto, S.L. and da Silva, E.A., 2020. A survey on performance metrics forobject-detection algorithms, In: Internat. Conf. Systems, Signals Image Process.(IWSSIP): 237-242.
  38. Petrov, I.B. and Khokhlov, N.I., 2014. Modeling 3D seismic problems using high-performance computing systems, Mathemat. Models Comput. Simulat., 6: 342-350.
  39. Pochet, A., Diniz, P.H., Lopes, H. and Gattass, M., 2018. Seismic fault detection usingconvolutional neural networks trained on synthetic poststacked amplitude maps.IEEE Geosci. Remote Sens. Lett., 16: 352-356.
  40. Ramsden, C., Bennett, G. and Long, A., 2005. High-resolution 3D seismic imaging inpractice. The Leading Edge, 24: 423-428.
  41. Redmon, J., Divvala, S., Girshick, R. and Farhadi, A., 2016. You only look once:
  42. Unified, real-time object detection. IEEE Conf. Comput. Vis. Patt. Recognit.(CVPR): 779-788.
  43. Redmon, J. and Farhadi, A., 2017. YOLO9000: better, faster, stronger. IEEE Conf.Comput. Vis. Patt. Recognit. (CVPR): 7263-7271.
  44. Redmon, J. and Farhadi, A., 2018. YOLOv3: An incremental improvement, arXivpreprint arXiv: 1804.02767.
  45. Robertsson, J.O., Blanch, J.O. and Symes, W.W., 1994. Viscoelastic finite-differencemodelling, Geophysics, 59: 1444-1456.
  46. Ryu, D., Kim, A. and Ha, W., 2015. Expanding domain method for 3D time-Laplace-domain hybrid modelling, Geosystem Eng., 18 (5), 259-265.
  47. Sadak, F., Saadat, M. and Hajiyavand, A.M., 2020. Real-time deep learning-based imagerecognition for applications in automated positioning and injection of biologicalcells, Comput. Biol. Med., 125: 103976.
  48. Santos, J.E., Douglas Jr., J., Morley, M.E. and Lovera, O.M., 1988. Finite elementmethods for a model for full waveform acoustic logging, IMA J. Numer. Anal., 8:415-433.
  49. Siahkoohi, A., Louboutin, M. and Herrmann, F.J., 2019. Neural network augmentedwave-equation simulation. arXiv preprint arXiv:1910.00925.
  50. Suh, S. and Wang, B., 2011. Expanding domain methods in GPU based TTI reverse timemigration. Expanded Abstr., 81st. Ann. Internat. SEG Mtg., San Antonio: 3460-
  51. Szegedy, C., Toshev, A. and Erhan, D., 2013. Deep neural networks for object detection.Advan. Neural Informat. Process. Syst., 1-9.
  52. Tang, X. and Shao, Q., 2013. Numerical simulation on seismic liquefaction by adaptivemesh refinement due to two recovered fields in error estimation. Soil Dynam.Earthq. Engineer., 49: 109-121.
  53. Vidale, J.E., 1990. Finite-difference calculation of traveltimes in three dimensions.Geophysics, 55: 521-526.
  54. Virieux, J., 1986. P-SV wave propagation in heterogeneous media: Velocity-stress finite-difference method. Geophysics, 51: 889-901.
  55. Vrolijk, J.W. and Blacquiére, G., 2021. Source deghosting of coarsely-sampled common-receiver data using a convolutional neural network. Geophysics, 86(3): V185-V196.
  56. Wang, B., Zhang, N., Lu, W., and Wang, J., 2019. Deep-learning-based seismic datainterpolation: A preliminary result. Geophysics, 84(1), V11-V20
  57. Womg, A., Shafiee, M.J., Li, F. and Chwyl, B., 2018. Tiny SSD: A tiny single-shotdetection deep convolutional neural network for real-time embedded objectdetection, 2018 15th Conference on Computer and Robot Vision (CRV), 95-101.
  58. Yang, F. and Ma, J., 2019. Deep-learning inversion: A next-generation seismic velocitymodel building method. Geophysics, 84(4), R583-R599.
Share
Back to top
Journal of Seismic Exploration, Print ISSN: 0963-0651, Published by AccScience Publishing
We use cookies on our website to ensure you get the best experience.
Read more about our cookies here
Accept