Análisis comparativo del uso de algoritmos matching pursuit para la migración sísmica sobre datos comprimidos
Resumen
Actualmente, la cantidad de datos recolectados en un estudio sísmico está en el orden de los cientos de Terabytes. Esta cantidad de datos impone varios retos computacionales, siendo uno de los más importantes el cuello de botella de entrada/salida datos desde/hacia la memoria principal y la memoria del nodo. Este cuello de botella es una consecuencia directa de que la velocidad de acceso a la memoria del disco es miles de veces más lenta que la velocidad de procesamiento de los coprocesadores (e.g. GPUs). Nosotros presentamos una migración especial de Kirchhoff que permite migrar datos comprimidos. Los datos sísmicos se comprimen utilizando tres algoritmos conocidos Matching Pursuit. Nuestro enfoque busca reducir el número de accesos de memoria requeridos por el operador de Kirchhoff. El enfoque requiere agregar más operaciones matemáticas a la migración tradicional de Kirchhoff. De esta manera, cambiamos las operaciones lentas (acceso a la memoria) por operaciones rápidas (operaciones matemáticas). Los resultados experimentales demuestran que el método propuesto
preserva significativamente los atributos sísmicos de la imagen para una relación de compresión de hasta 20:1.
Referencias bibliográficas
Wu, C., T. Wang, H. Wang, H. Li, and S. Liu, (2018). Full 3D double square root migration method for VTI media and its application in real data, Int. Geophys. Conf. Beijing, China, 24-27 April 2018, 600-603. https://doi.org/10.1190/IGC2018-147
Gregg, C. and K. Hazelwood, (2011). Where is the data? Why you cannot debate CPU vs. GPU performance without the answer, ISPASS 2011 - IEEE Int. Symp. Perform. Anal. Syst. Softw., 134-144. https://doi.org/10.1109/ISPASS.2011.5762730
Salamanca, W. A., A.-B. Ramirez, and F.-A. Vivas, (2018). Comparative analysis of 3D RTM Implementation Strategies for an Efficient Use of Memory in a Single GPU, CT&F-Ciencia, Tecnol. y Futur., 8(2), 75-82. https://doi.org/10.29047/01225383.83
Fajardo, C. A., Ó. M. Reyes, and J. (Universidad R. J. C. Castillo, (2018). Reducing the I/O Bottleneck by a Compression Strategy, Eng. Lett., 26(2), 203-209. [Online]. Available: http://www.engineeringletters.com/issues_v26/issue_2/EL_26_2_01.pdf
Bouska, J. and S. Gray, (1998). Migration of unequally sampled compressed seismic data, Expand. Abstr. Tech. Program, SEG 68th Annu. Meet., 1128-1130. https://doi.org/10.1190/1.1820087
Yu, Z., G. A. McMechan, P. D. Anno, and J. F. Ferguson, (2004). Wavelet-transform-based prestack multiscale Kirchhoff migration, Geophysics, 69(6), 1505-1512. https://doi.org/10.1190/1.1836823
Zheludev, V. A., E. Ragoza, and D. D. Kosloff, (2002). Fast Kirchhoff migration in the wavelet domain, Explor. Geophys., 33(1), 23-27. https://doi.org/10.1071/EG02023
Alkhalifah, T., (2011). Efficient traveltime compression for 3D prestack Kirchhoff migration, Geophys. Prospect., 59(1), 1-9. https://doi.org/10.1111/j.1365-2478.2010.00886.x
Mallat, S. G., (1993). Matching pursuits with time-frequency dictionaries, IEEE Trans. Signal Process., 41(12), 3397-3415. https://doi.org/10.1109/78.258082
Wang, B. and K. Pann, (1996). Kirchhoff migration of seismic data compressed by matching pursuit decomposition, in Expanded abstracts of the technical program, SEG 66th annual meeting, (1), 1642--1645. https://doi.org/10.1190/1.1826441
Li, X.--G., B. Wang, K. Pann, J. Anderson, and L. Deng, (1998). Fast migration using a matching pursuit algorithm, in Expanded Abstracts of the Technical Program, SEG 68th Annual Meeting, 1732-1735. https://doi.org/10.1190/1.1820261
Lin, L., B. Shi, and P. An, (2016). Multiwavelet prestack Kirchhoff migration, Geophysics, 81(3), S79-S85. https://doi.org/10.1190/geo2015-0140.1
Xiongwen, W., W. Huazhong, and Z. Xiaopeng, (2014). High dimensional seismic data interpolation with weighted matching pursuit based on compressed sensing, J. Geophys. Eng., 11(6). https://doi.org/10.1088/1742-2132/11/6/065003
Wang, Y., (2007). Seismic time-frequency spectral decomposition by matching pursuit, Geophysics, 72(1), V13-V20. https://doi.org/10.1190/1.2387109
Liu, J. and K. J. Marfurt, (2005). Matching pursuit using Morlet wavelets., in 2005 SEG Annual Meeting, (4), 786-790. https://doi.org/10.1190/1.2148276
Lin, H., Y. Li, H. Ma, B. Yang, and J. Dai, (2015). Matching-pursuit-based spatial-trace time-frequency peak filtering for seismic random noise attenuation, IEEE Geosci. Remote Sens. Lett., 12(2), 394-398. https://doi.org/10.1109/LGRS.2014.2344020
Gholamy, A. and V. Kreinovich, (2014). Why Ricker wavelets are successful in processing seismic data: Towards a theoretical explanation, IEEE SSCI 2014 - 2014 IEEE Symp. Ser. Comput. Intell. - CIES 2014 2014 IEEE Symp. Comput. Intell. Eng. Solut. Proc., 11-16. https://doi.org/10.1109/CIES.2014.7011824
Duval, L. C. and V. Buitran, (2001). Compression denoising: using seismic compression for uncoherent noise removal, EAGE 63rd Conf. Tech. Exhib. Netherlands, (June). https://doi.org/10.3997/2214-4609-pdb.15.A-21
Cai, T. T. and L. Wang, (2011). Orthogonal matching pursuit for sparse signal recovery with noise, IEEE Trans. Inf. Theory, 57(7), 4680-4688. https://doi.org/10.1109/TIT.2011.2146090
Pati, Y. C., R. Rezaiifar, and P. S. Krishnaprasad, (1993). Orthogonal matching pursuit: Recursive function approximation with applications to wavelet decomposition, in Signals, Systems and Computers, 1993. 1993 Conference Record of The Twenty-Seventh Asilomar Conference on, 40-44. [Online]. Available: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.348.5735
Santos*, P. N. and R. C. Pestana, (2015). Least-squares Kirchhoff migration using traveltimes based on the maximum amplitude criterion by the rapid expansion method, in 14th International Congress of the Brazilian Geophysical Society & EXPOGEF, Rio de Janeiro, Brazil, 3-6 August 2015, 1043-1047. https://doi.org/10.1190/sbgf2015-207
Vidale, J., (1988). Finite-difference calculation of travel times, Bull. Seismol. Soc. Am., 78(6), 2062-2076. Available at: https://earthweb.ess.washington.edu/vidale/Reprints/BSSA/1988_Vidale.pdf
Descargas
Datos de los fondos
-
Departamento Administrativo de Ciencia, Tecnología e Innovación (COLCIENCIAS)
Números de la subvención project grant 0266 of 2013 -
Ecopetrol
Números de la subvención 0266 of 2013
