심층 학습 기법을 이용한 탄성파 자료 잡음 제거 적용사례 분석 = Case Analysis of Applications of Seismic Data Denoising Methods using Deep-Learning Techniques

다국어 초록 (Multilingual Abstract)

Recent rapid advances in computer hardware performance have led to relatively low computational costs, increasing the number of applications of machine-learning techniques to geophysical problems. In particular, deep-learning techniques are gaining in popularity as the number of cases successfully solving complex and nonlinear problems has gradually increased. In this paper, applications of seismic data denoising methods using deep-learning techniques are introduced and investigated. Depending on the type of attenuated noise, these studies are grouped into denoising applications of coherent noise, random noise, and the combination of these two types of noise. Then, we investigate the deep-learning techniques used to remove the corresponding noise. Unlike conventional methods used to attenuate seismic noise, deep neural networks, a typical deep-learning technique, learn the characteristics of the noise independently and then automatically optimize the parameters. Therefore, such methods are less sensitive to generalized problems than conventional methods and can reduce labor costs. Several studies have also demonstrated that deep-learning techniques perform well in terms of computational cost and denoising performance. Based on the results of the applications covered in this paper, the pros and cons of the deep-learning techniques used to remove seismic noise are analyzed and discussed.

국문 초록 (Abstract)

최근 컴퓨터 하드웨어 성능의 급속한 발전으로 인해 계산 비용이 상대적으로 낮아지면서 기계 학습 기법을 지구물리학적 문제에 적용하는 사례가 점차 증가하고 있다. 특히 심층 학습 기법이 복잡하고 비선형적인 문제를 성공적으로해결하는 사례가 많아지면서 큰 인기를 얻고 있다. 이 논문에서는 심층 학습 기법을 이용한 탄성파 자료 잡음 제거 적용사례를 조사하고 소개하였다. 감쇠하고자 하는 잡음 유형에 따라 일관성 잡음 적용사례, 무작위 잡음 적용사례, 일관성잡음 및 무작위 잡음 적용사례로 분류하였고 해당 잡음 제거에 사용된 심층 학습 기법에 대해 조사하였다. 대표적인 심층 학습 기법인 심층 신경망은 탄성파 잡음 제거에 사용된 기존 기법과 달리 잡음의 특징을 스스로 학습하며 매개변수를 자동으로 최적화한다. 따라서 기존 기법에 비해 일반화 문제에 덜 민감하며 인적 비용을 절감할 수 있다. 또한 여러연구 사례를 통해 계산 비용이나 잡음 제거 성능 측면에서도 심층 학습 기법이 뛰어난 성과를 달성하는 것을 보여주었다. 연구 결과들을 토대로 탄성파 잡음 제거에 사용된 심층 학습 기법의 장단점에 대해 분석하고 논의하였다.

참고문헌 (Reference)

1 Zheng, Y., "The improved DnCNN for linear noise attenuation" 56-59, 2020

2 Xing Zhao, "Swell-noise attenuation: A deep learning approach" Society of Exploration Geophysicists 38 (38): 934-942, 2019

3 Vicente Oropeza, "Simultaneous seismic data denoising and reconstruction via multichannel singular spectrum analysis" Society of Exploration Geophysicists 76 (76): V25-V32, 2011

4 Jia, Z., "Separating ground-roll from land seismic record via convolutional neural network" 60-63, 2018

5 Kim, Y., "Seismic random noise attenuation in f-x domain using complex-valued residual convolutional neural network" 2579-2583, 2019

6 Jin, Y., "Seismic data denoising by deep-residual networks" 4593-4597, 2018

7 Yilmaz, Ö., "Seismic data analysis: processing, inversion and interpretation of seismic data, Vol. I" Soc. Expl. Geophys 837-998, 2001

8 Liu, D., "Random-noise suppression in seismic data: What can deep learning do?" 2016-2020, 2018

9 Canales, L. L., "Random noise reduction" 525-527, 1984

10 Zhang, F., "Random noise attenuation method for seismic data based on deep residual networks" 1774-1777, 2018

11 Si, X., "Random noise attenuation based on residual learning of deep convolutional neural network" 1986-1990, 2018

12 Ledig, C., "Photo-realistic single image super-resolution using a generative adversarial network" 4681-4690, 2017

13 Zhang, Y., "Noise attenuation for seismic image using a deep-residual learning" 2176-2180, 2018

14 H.A. Rowley, "Neural network-based face detection" Institute of Electrical and Electronics Engineers (IEEE) 20 (20): 23-38, 1998

15 Liu, D., "Must we have labels for denoising seismic data based on deep learning?" 31-35, 2020

16 Zhang, C., "Microseismic and seismic noise attenuation by supervised deep learning with noisy natural images" 4485-4489, 2019

17 Ray Abma, "Lateral prediction for noise attenuation by t-x and f-x techniques" Society of Exploration Geophysicists 60 (60): 1887-1896, 1995

18 Xu, J., "Joint dictionary learning and super-resolution network on seismic noise removal" 2473-2477, 2019

19 Hayit Greenspan, "Guest Editorial Deep Learning in Medical Imaging: Overview and Future Promise of an Exciting New Technique" Institute of Electrical and Electronics Engineers (IEEE) 35 (35): 1153-1159, 2016

20 Andrew J. Deighan, "Ground‐roll suppression using the wavelet transform" Society of Exploration Geophysicists 62 (62): 1896-1903, 1997

21 Li, H., "Ground-roll noise attenuation based on convolutional neural network" 69-73, 2019

22 Si, X., "Ground roll attenuation based on conditional and cycle generative adversarial networks" 95-98, 2020

23 Xie, P., "Generativeadversarial network-based fast-noise removal on land-seismic data" 2171-2175, 2018

24 Goodfellow, I., "Generative adversarial nets" 2672-2680, 2014

25 Trickett, S., "F-xy cadzow noise suppression" 2586-2590, 2008

26 Chen, Y., "Dip-separated structural filtering using seislet transform and adaptive empirical mode decomposition based dip filter" 206 (206): 457-469, 2016

27 He, K., "Deep residual learning for image recognition" 770-778, 2016

28 Li, H., "Deep learning for groundroll noise attenuation" 1981-1985, 2018

29 Siwei Yu, "Deep learning for denoising" Society of Exploration Geophysicists 84 (84): V333-V350, 2019

30 Goodfellow, I., "Deep learning" MIT Press 2016

31 Chollet, F., "Deep Learning with Python" Manning 29-35, 2018

32 Kai Zhang, "Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising" Institute of Electrical and Electronics Engineers (IEEE) 26 (26): 3142-3155, 2017

33 Yarham, C., "Bayesian ground-roll separation by curvelet-domain sparsity promotion" 2576-2580, 2008

34 Enning Wang, "Applying machine learning to 3D seismic image denoising and enhancement" Society of Exploration Geophysicists 7 (7): SE131-SE139, 2019

35 Liu, J., "3-d high resolution radon transforms applied to ground-roll suppression in orthogonal seismic surveys" 2144-2147, 2004