Journal of Korean Tunnelling and Underground Space Association (한국터널지하공간학회 논문집)

Korean Tunnelling and Underground Space Association (한국터널지하공간학회)
권호 표지
DOI QR코드

DOI QR Code

Crack detection in concrete using deep learning for underground facility safety inspection

지하시설물 안전점검을 위한 딥러닝 기반 콘크리트 균열 검출

  • Eui-Ik Jeon (AI Team, Innopam Co., Ltd.) ;
  • Impyeong Lee (Dept. of Geoinformatics, University of Seoul) ;
  • Donggyou Kim (Dept. of Geotechnical Engineering Research, Korea Institute of Civil Engineering and Building Technology)
  • 전의익 ((주)이노팸 인공지능팀) ;
  • 이임평 (서울시립대학교 공간정보공학과) ;
  • 김동규 (한국건설기술연구원 지반연구본부)
  • Received : 2023.10.31
  • Accepted : 2023.11.14
  • Published : 2023.11.30
Download PDF
⟨ Previous Next ⟩

Abstract

The cracks in the tunnel are currently determined through visual inspections conducted by inspectors based on images acquired using tunnel imaging acquisition systems. This labor-intensive approach, relying on inspectors, has inherent limitations as it is subject to their subjective judgments. Recently research efforts have actively explored the use of deep learning to automatically detect tunnel cracks. However, most studies utilize public datasets or lack sufficient objectivity in the analysis process, making it challenging to apply them effectively in practical operations. In this study, we selected test datasets consisting of images in the same format as those obtained from the actual inspection system to perform an objective evaluation of deep learning models. Additionally, we introduced ensemble techniques to complement the strengths and weaknesses of the deep learning models, thereby improving the accuracy of crack detection. As a result, we achieved high recall rates of 80%, 88%, and 89% for cracks with sizes of 0.2 mm, 0.3 mm, and 0.5 mm, respectively, in the test images. In addition, the crack detection result of deep learning included numerous cracks that the inspector could not find. if cracks are detected with sufficient accuracy in a more objective evaluation by selecting images from other tunnels that were not used in this study, it is judged that deep learning will be able to be introduced to facility safety inspection.

현재 지하시설물의 균열을 영상 취득 시스템으로 취득한 경우 점검자가 취득된 영상에서 육안검사를 수행하여 미세균열을 판단한다. 점검자에 의존한 노동집약적인 방법은 점검자의 주관적인 판단에 영향을 받는 문제점을 가지고 있다. 최근에는 딥러닝을 활용하여 자동으로 콘크리트 균열을 탐지하기 위한 연구가 활발하게 수행되고 있다. 대부분의 연구에서는 공개 데이터셋을 활용하거나 분석과정의 객관성이 충분하지 못해 실제 업무에 적용하기 어려운 점이 있다. 본 연구는 실제 검사 시스템과 동일한 형태의 영상을 시험 데이터셋으로 선정하여 딥러닝 모델들을 평가하였다. 균열 탐지의 정확도를 향상시키기 위하여 딥러닝 모델들의 장단점을 상호 보완할 수 있는 앙상블 기법을 적용하였다. 시험 영상에서 폭 0.2 mm, 0.3 mm 및 0.5 mm의 균열들은 각각 80%, 88% 및 89%의 높은 재현율로 탐지되었다. 딥러닝을 적용한 균열 탐지 결과에서는 점검자의 육안 검수 과정에 찾지 못한 다수의 균열들을 포함하고 있었다. 향후 본 연구에서 사용하지 않은 다른 터널의 영상을 시험 영상으로 선정하여 보다 더 객관적인 평가에서 충분한 정확도로 균열을 탐지하게 된다면, 시설물 안점 점검 방식에 딥러닝의 도입이 가능할 것으로 판단된다.

Keywords

Acknowledgement

본 연구는 한국건설기술연구원 주요사업으로 지원을 받아 수행된 연구(인공지능을 활용한 대심도 지하 대공간의 스마트 복합 솔루션 개발)로 이에 감사합니다.

References

  1. Bae, S.H. (2023), Deep learning-based crack detection for facility safety inspection - focusing on tunnel cement concrete lining, Master Thesis, University of Seoul, pp. 35-46.
  2. Bae, S.H., Ham, S.W., Lee, I.P., Lee, G.P., Kim, D.G. (2022), "Deep learning based crack detection from tunnel cement concrete lining", Journal of Korean Tunnelling and Underground Space Association, Vol. 24, No. 6, pp. 583-598. https://doi.org/10.9711/KTAJ.2022.24.6.583
  3. Chen, L.C., Zhu, Y., Papandreou, G., Schroff, F., Adam, H. (2018), "Encoder-decoder with atrous separable convolution for semantic image segmentation", Proceedings of the European Conference on Computer Vision (ECCV), Munich, Germany, pp. 801-818.
  4. Cheng, B., Misra, I., Schwing, A.G., Kirillov, A., Girdhar, R. (2022), "Masked-attention mask transformer for universal image segmentation", Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, USA, pp. 1290-1299.
  5. Hadinata, P.N., Simanta, D., Eddy, L., Nagai, K. (2021), "Crack detection on concrete surfaces using deep encoder-decoder convolutional neural network: a comparison study between U-Net and DeepLabV3+", Journal of the Civil Engineering Forum, Vol. 7, No. 3, pp. 323-334. https://doi.org/10.22146/jcef.65288
  6. Ham, S.W., Bae, S.H., Lee, I.P., Lee, G.P., Kim, D.G. (2022), "An evaluation methodology for cement concrete lining crack segmentation deep learning model", Journal of Korean Tunnelling and Underground Space Association, Vol. 24, No. 6, pp. 513-524.
  7. Jeon, E.I., Kim, S.H., Park, S.Y., Kwak, J.W., Choi, I.H. (2021), "Semantic segmentation of seagrass habitat from drone imagery based on deep learning: a comparative study", Ecological Informatics, Vol. 66, 101430.
  8. Kim, A.R., Kim, D., Byun, Y.S., Lee, S.W. (2018), "Crack detection of concrete structure using deep learning and image processing method in geotechnical engineering", Journal of the Korean Geotechnical Society, Vol. 34, No. 12, pp. 145-154.
  9. Kim, B.H., Cho, S.J. (2018), "Automated vision-based detection of cracks on concrete surfaces using a deep learning technique", Sensors, Vol. 18, No. 10, pp. 3452.
  10. Kim, H.S., Kim, M.H. (2022), "A study on the image-based malware classification system that combines image preprocessing and ensemble techniques for high accuracy", KIPS Transactions on Computer and Communication Systems, Vol. 11, No. 7, pp. 225-232.
  11. Lee, T.H., Kim, J.H., Lee, S.J., Ryu, S.K., Joo, B.C. (2023), "Improvement of concrete crack segmentation performance using stacking ensemble learning", Applied Sciences, Vol. 13, No. 4, 2367.
  12. Liu, Y., Yao, J., Lu, X., Xie, R., Li, L. (2019), "DeepCrack: A deep hierarchical feature learning architecture for crack segmentation", Neurocomputing, Vol. 338, pp. 139-153. https://doi.org/10.1016/j.neucom.2019.01.036
  13. Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., Guo, B. (2021), "Swin transformer: Hierarchical vision transformer using shifted windows", Proceedings of the IEEE/CVF International Conference on Computer Vision, Montrealers, Canada, pp. 10012-10022.
  14. Pandey, R.K., Achara, A. (2022), CoreDeep: Improving crack detection algorithms using width stochasticity, arXiv preprint arXiv:2209.04648.
  15. Wang, W., Dai, J., Chen, Z., Huang, Z., Li, Z., Zhu, X., Hu, X., Lu, T., Lu, L., Li, H., Wang, X., Qiao, Y. (2023), "Internimage: Exploring large-scale vision foundation models with deformable convolutions", Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, Canada, pp. 14408-14419.
  16. Xiao, T., Liu, Y., Zhou, B., Jiang, Y., Sun, J. (2018), "Unified perceptual parsing for scene understanding", Proceedings of the European Conference on Computer Vision (ECCV), Munich, Germany, pp. 418-434.