All Issue

2025 Vol.24, Issue 1 Preview Page

Research Article

Advanced Rock Type Classification Using Deep Learning: A Data-Driven Approach to Tunnel Stability Monitoring and Maintenance

딥러닝을 활용한 암석 분류 프레임워크: 터널 안정성 모니터링 및 유지관리를 위한 데이터 기반 접근 방식

Seungjoo Lee1
,
Yongseong Kim2
,
Yongjin Kim3
,
Jongseol Park4
,
Bongjun Ji5*
이 승주1
,
김 용성2
,
김 용진3
,
박 종설4
,
지 봉준5*
1Member, Researcher, Korean Peninsula Infrastructure Research Center, Department of Construction Policy Research, Korea Institute of Civil Engineering and Building Technology, 283 Goyang-Daero, Ilsanseo-Gu, Goyang-Si, Gyeonggi-Do 10223, Republic of Korea
2Member, Professor, Department of Regional Infrastructure Engineering, Kangwon National University, 1 Kangwondaehak-gil, Chuncheon-si, Gangwon-do 24341, Republic of Korea
3Member, CEO, Smart E&C, Kangwondaehak-gil, Chuncheon-si, Gangwon-do 24341, Republic of Korea
4Non-Member, Executive Director, Smart E&C, Kangwondaehak-gil, Chuncheon-si, Gangwon-do 24341, Republic of Korea
5Member, Assistant Professor, Graduate School of Data Science, Pusan National University, 2 Busandaehak-ro 63beon-gil, Geumjeong-gu, Busan 46241, Republic of Korea
*Corresponding Author
30 March 2025. pp. 89-100
PDF XML Citation
Abstract

Tunnels serve a pivotal role in various infrastructure systems—including roads, railways, and underground facilities—and ensuring their stability is of critical economic and societal importance. Traditional rock classification methods often rely on on-site geological surveys and petrographic/chemical analyses, which can be time-consuming, cost-intensive, and prone to subjective judgments. With the recent advancement in big data and artificial intelligence (AI) technologies, particularly Deep Learning, it has become possible to automatically learn and classify intricate rock patterns from large-scale image datasets. In this study, we propose a deep learning-based rock classification framework and apply several state-of-the-art models—ResNet, EfficientNet, MobileNet, and others—to both broad categories (igneous, sedimentary, metamorphic) and more fine-grained rock types (e.g., marble, sandstone, phyllite). Our results show that while high accuracies exceeding 90% can be achieved for three-class problems, classification performance drops significantly when expanded to 18 or more subclasses, likely due to the complex microstructures and ambiguous boundaries among rock types. These findings highlight the substantial potential of deep learning in tunnel stability monitoring and maintenance, yet also underscore the necessity of hierarchical classification strategies, multimodal data integration (e.g., microscopic or hyperspectral imaging), and further hyperparameter optimization for more refined rock classification tasks.

Keywords
Deep learning, Rock classification
Tunnel stability
Geotechnical engineering
Machine learning
Maintenance

터널은 도로, 철도, 지하 인프라 등 다양한 산업 분야에서 핵심적인 역할을 수행하며, 터널 안정성 확보는 경제적·사회적으로 매우 중요한 과제이다. 이러한 터널의 안정석 확보를 위해서는 암반에 대한 이해가 필수적이며 이를 위해 기존에는 현장 지질조사와 광물학적·지구화학적 분석 등을 통해 암석을 분류해왔으나, 이는 시간·비용 측면에서 비효율적이고 경험적 판단에 의존하기 쉬운 한계가 있었다. 최근 빅데이터와 인공지능(AI) 기술, 특히 딥러닝(Deep learning)의 급속한 발전으로 인해, 대규모 암석 이미지를 통해 복잡한 패턴을 자동으로 학습하고 분류할 수 있는 가능성이 열리고 있다. 본 연구에서는 딥러닝을 활용한 암반 분류 프레임워크를 제안하고, 화성암·퇴적암·변성암을 비롯하여 여러 세분류 암석(예: 대리암, 사암, 편암 등)에 대해 다양한 딥러닝 모델을 적용·비교·평가하였다. 연구 결과, 대분류(3분류) 단계에서는 ResNet, EfficientNet, MobileNet 등에서 90% 이상의 높은 정확도를 달성하였으나, 18개 이상의 세분류로 확대할 경우 전반적으로 정확도가 급격히 떨어지는 현상이 관찰되었다. 이는 암석의 조직과 광물 조합이 매우 복잡하고, 클래스 간 경계가 모호해지는 점이 주요 원인으로 분석된다. 본 연구를 통해, 딥러닝 기반 암석 분류가 터널 안정성 모니터링 및 유지관리 분야에서 높은 잠재력을 지님을 확인하였으나, 세분화된 암석 분류 정확도를 높이기 위해서는 추가 데이터가 요구됨을 확인할 수 있었다.

References
1

Baghbani, A., Choudhury, T., Costa, S. and Reiner, J. (2022), "Application of Artificial Intelligence in Geotechnical Engineering: A State-of-the-Art Review", Earth-Science Reviews, Vol.228, 103991.

10.1016/j.earscirev.2022.103991
2

Carpenter, C. (2020), "Machine-Learning Image Recognition Enhances Rock Classification", Journal of Petroleum Technology, Vol.72, No.10, pp.63-64.

10.2118/1020-0063-JPT
3

Chen, J., Yang, T., Zhang, D., Huang, H. and Tian, Y. (2021), "Deep Learning Based Classification of Rock Structure of Tunnel Face", Geoscience Frontiers, Vol.12, No.1, pp.395-404.

10.1016/j.gsf.2020.04.003
4

De Lima, R. P., Bonar, A., Coronado, D. D., Marfurt, K. and Nicholson, C. (2019), "Deep Convolutional Neural Networks as a Geological Image Classification Tool", The Sedimentary Record, Vol.17, No.2, pp.4-9.

10.2110/sedred.2019.2.4
5

Fernandez-Gutierrez, J. D., Rodriguez, S. S., Gonzalo-Orden, H. and Perez-Acebo, H. (2021), "Analysis of Rock Mass Classifications for Safer Infrastructures", Transportation Research Procedia, Vol.58, pp.606-613.

10.1016/j.trpro.2021.11.080
6

Hasegawa, N., Hasegawa, S., Kitaoka, T. and Ohtsu, H. (2019), "Applicability of Neural Network in Rock Classification of Mountain Tunnel", Materials Transactions, Vol.60, No.5, pp.758-764.

10.2320/matertrans.Z-M2019809
7

He, K., Zhang, X., Ren, S. and Sun, J. (2016), "Deep Residual Learning for Image Recognition", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.770-778.

10.1109/CVPR.2016.90
8

Howard, A. G. (2017), "MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications", arXiv preprint, arXiv:1704.04861.

9

Kim, N., Sim, J., Won, J. P. and Shim, J. W. (2005), "An Experimental Study of Polypropylene Fiber for the Prevention of Explosive Spalling of Tunnel Concrete Lining", Journal of Korean Tunnelling and Underground Space Association, Vol.7, No.4, pp.323-333.

10

Krizhevsky, A., Sutskever, I. and Hinton, G. E. (2012), "ImageNet Classification with Deep Convolutional Neural Networks", Advances in Neural Information Processing Systems, Vol.25.

11

LeCun, Y., Bottou, L., Bengio, Y. and Haffner, P. (1998), "Gradient-Based Learning Applied to Document Recognition", Proceedings of the IEEE, Vol.86, No.11, pp.2278-2324.

10.1109/5.726791
12

Liu, W., Li, A., Fang, W., Love, P. E., Hartmann, T. and Luo, H. (2023), "A Hybrid Data-Driven Model for Geotechnical Reliability Analysis", Reliability Engineering & System Safety, Vol.231, 108985.

10.1016/j.ress.2022.108985
13

Młynarczuk, M., Górszczyk, A. and Ślipek, B. (2013), "The Application of Pattern Recognition in the Automatic Classification of Microscopic Rock Images", Computers & Geosciences, Vol.60, pp.126-133.

10.1016/j.cageo.2013.07.015
14

Moon, B. C., Jeong, J. and Chung, C. H. (2005), "The Classifying Ability of the Igneous Rocks with Naked Eyes for Preservice Science Teachers", Journal of the Korean Earth Science Society, Vol.26, No.7, pp.630-639.

15

Ng, A. Y. (2004), "Feature Selection, L1 vs. L2 Regularization, and Rotational Invariance", Proceedings of the Twenty-First International Conference on Machine Learning, p.78.

10.1145/1015330.1015435
16

Ran, X., Xue, L., Zhang, Y., Liu, Z., Sang, X. and He, J. (2019), "Rock Classification from Field Image Patches Analyzed Using a Deep Convolutional Neural Network", Mathematics, Vol.7, No.8, 755.

10.3390/math7080755
17

Sandler, M., Howard, A., Zhu, M., Zhmoginov, A. and Chen, L. (2018), "MobileNetV2: Inverted Residuals and Linear Bottlenecks", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.4510-4520.

10.1109/CVPR.2018.00474
18

Sun, Y. and Gu, Z. (2022), "Using Computer Vision to Recognize Construction Material: A Trustworthy Dataset Perspective", Resources, Conservation and Recycling, Vol.183, 106362.

10.1016/j.resconrec.2022.106362
19

Tan, M. and Le, Q. (2019), "EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks", Proceedings of the International Conference on Machine Learning, PMLR, pp.6105-6114.

Information
  • Publisher :Korean Geosythetics Society
  • Publisher(Ko) :한국지반신소재학회
  • Journal Title :Journal of the Korean Geosynthetics Society
  • Journal Title(Ko) :한국지반신소재학회 논문집
  • Volume : 24
  • No :1
  • Pages :89-100
  • Received Date : 2025-01-08
  • Revised Date : 2025-02-16
  • Accepted Date : 2025-03-12
  • DOI :https://doi.org/10.12814/jkgss.2025.24.1.089
Journal Informaiton Journal of the Korean Geosynthetics Society Journal of the Korean Geosynthetics Society
  • NRF
  • KOFST
  • KISTI Current Status
  • KISTI Cited-by
  • crosscheck
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close