Structural Engineering and Mechanics

Techno-Press (테크노프레스)
권호 표지
DOI QR코드

DOI QR Code

Transfer learning for crack detection in concrete structures: Evaluation of four models

  • Received : 2024.03.14
  • Accepted : 2024.07.02
  • Published : 2024.07.25
⟨ Previous Next ⟩

Abstract

The objective of this research is to improve public safety in civil engineering by recognizing fractures in concrete structures quickly and correctly. The study offers a new crack detection method based on advanced image processing and machine learning techniques, specifically transfer learning with convolutional neural networks (CNNs). Four pre-trained models (VGG16, AlexNet, ResNet18, and DenseNet161) were fine-tuned to detect fractures in concrete surfaces. These models constantly produced accuracy rates greater than 80%, showing their ability to automate fracture identification and potentially reduce structural failure costs. Furthermore, the study expands its scope beyond crack detection to identify concrete health, using a dataset with a wide range of surface defects and anomalies including cracks. Notably, using VGG16, which was chosen as the most effective network architecture from the first phase, the study achieves excellent accuracy in classifying concrete health, demonstrating the model's satisfactorily performance even in more complex scenarios.

Keywords

Acknowledgement

The authors acknowledge that no financial support was received for this research.

References

  1. Bastug, E., Bennis, M. and Debbah, M. (2015), "A transfer learning approach for cache-enabled wireless networks", 13th International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOpt), US and Canada.
  2. Brooks, R.A. (1991), "New approaches to robotics", Sci., 253(5025), 1227-1232. https://doi.org/10.1126/science.253.5025.1227.
  3. Cote-Allard, U., Fall, C.L., Drouin, A., Campeau-Lecours, A., Gosselin, C., Glette, K., Laviolette, F. and Gosselin, B. (2019), "Deep learning for electromyographic hand gesture signal classification using transfer learning", IEEE Trans. Neur. Syst. Rehab. Eng., 27(4), 760-771. https://doi.org/10.1109/TNSRE.2019.2896269.
  4. Deng, J., Zhang, Z., Marchi, E. and Schuller, B. (2013), "Sparse autoencoder-based feature transfer learning for speech emotion recognition", Humaine Association Conference on Affective Computing and Intelligent Interaction, Geneva, Switzerland.
  5. Deshpande, A. (2016), A Beginner's Guide to Understanding Convolutional Neural Networks.
  6. Dorafshan, S., Maguire, M. and Qi, X. (2016), Automatic Surface Crack Detection in Concrete Structures using OTSU Thresholding and Morphological Operations, Civil and Environmental Engineering Faculty Publications.
  7. Dorafshan, S., Thomas, R.J. and Maguire, M. (2018), "SDNET2018: An annotated image dataset for non-contact concrete crack detection using deep convolutional neural networks", Data Brief, 21, 1664-1668. https://doi.org/10.1016/j.dib.2018.11.015.
  8. Dudani, S.A. (1976), "The distance-weighted k-nearest-neighbor rule", IEEE Trans. Syst. Man Cybernet., SMC-6(4), 325-327. https://doi.org/10.1109/TSMC.1976.5408784.
  9. Durairaj, R., Varatharajan, T., Srinivasan, S.K., Gurupatham, B.G.A. and Roy, K. (2023), "Experimental investigation on flexural behaviour of sustainable reinforced concrete beam with a smart mortar layer", J. Compos. Sci., 7(4), 132. https://doi.org/10.3390/jcs7040132.
  10. Ellenberg, A., Kontsos, A., Moon, F. and Bartoli, I. (2016), "Bridge related damage quantification using unmanned aerial vehicle imagery", Struct. Control Hlth. Monit., 23(9), 1168-1179. https://doi.org/10.1002/stc.1831.
  11. Gelenbe, E. (1990), "Stability of the random neural network model", Neur. Comput., 2(2), 239-247. https://doi.org/10.1162/neco.1990.2.2.239.
  12. German, S., Brilakis, I. and DesRoches, R. (2012), "Rapid entropy-based detection and properties measurement of concrete spalling with machine vision for post-earthquake safety assessments", Adv. Eng. Inform., 26(4), 846-858. https://doi.org/10.1016/j.aei.2012.06.005.
  13. Gopalakrishnan, K., Khaitan, S.K., Choudhary, A. and Agrawal, A. (2017), "Deep convolutional neural networks with transfer learning for computer vision-based data-driven pavement distress detection", Constr. Build. Mater., 157, 322-330. https://doi.org/10.1016/j.conbuildmat.2017.09.110.
  14. Hamrat, M., Boulekbache, B., Chemrouk, M. and Amziane, S. (2016), "Flexural cracking behavior of normal strength, high strength and high strength fiber concrete beams, using Digital Image Correlation technique", Constr. Build. Mater., 106, 678-692. https://doi.org/10.1016/j.conbuildmat.2015.12.166.
  15. 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.
  16. Hosmer Jr, D.W., Lemeshow, S. and Sturdivant, R.X. (2013), Applied Logistic Regression, John Wiley & Sons.
  17. Huang, G., Liu, Z., Van Der Maaten, L. and Weinberger, K.Q. (2017), "Densely connected convolutional networks", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.
  18. Huthwohl, P. and Brilakis, I. (2018), "Detecting healthy concrete surfaces", Adv. Eng. Inform., 37, 150-162. https://doi.org/10.1016/j.aei.2018.05.004.
  19. Hwang, T. and Kuang, R. (2010), "A heterogeneous label propagation algorithm for disease gene discovery", Proceedings of the 2010 SIAM International Conference on Data Mining.
  20. Jahanshahi, M.R. and Masri, S.F. (2012), "Adaptive vision-based crack detection using 3D scene reconstruction for condition assessment of structures", Auto. Constr., 22, 567-576. https://doi.org/10.1016/j.autcon.2011.11.018.
  21. Jerez, J.M., Molina, I., Garcia-Laencina, P.J., Alba, E., Ribelles, N., Martin, M. and Franco, L. (2010), "Missing data imputation using statistical and machine learning methods in a real breast cancer problem", Artif. Intel. Medic., 50(2), 105-115. https://doi.org/10.1016/j.artmed.2010.05.002.
  22. Kanagaraj, B., Kiran, T., Gunasekaran, J., Nammalvar, A., Arulraj, P., Gurupatham, B.G.A. and Roy, K. (2022), "Performance of sustainable insulated wall panels with geopolymer concrete", Mater., 15(24), 8801. https://doi.org/10.3390/ma15248801.
  23. Kanagaraj, B., Nammalvar, A., Andrushia, A.D., Gurupatham, B.G.A. and Roy, K. (2023), "Influence of nano composites on the impact resistance of concrete at elevated temperatures", Fire, 6(4), 135. https://doi.org/10.3390/fire6040135.
  24. Kim, H., Ahn, E., Cho, S., Shin, M. and Sim, S.H. (2017), "Comparative analysis of image binarization methods for crack identification in concrete structures", Cement Concrete Res., 99, 53-61. https://doi.org/10.1016/j.cemconres.2017.04.018.
  25. Krizhevsky, A., Sutskever, I. and Hinton, G.E. (2017), "ImageNet classification with deep convolutional neural networks", Commun. ACM, 60(6), 84-90.
  26. Lawrence, S., Burns, I., Back, A., Tsoi, A.C. and Giles, C.L. (2002), Neural Network Classification and Prior Class Probabilities, Springer
  27. Lewis, R.J. (2000), "An introduction to classification and regression tree (CART) analysis", Annual Meeting of the Society for Academic Emergency Medicine in San Francisco, California.
  28. Li, L., Wang, Q., Zhang, G., Shi, L., Dong, J. and Jia, P. (2018), "A method of detecting the cracks of concrete undergo hightemperature", Constr. Build. Mater., 162, 345-358. https://doi.org/10.1016/j.conbuildmat.2017.12.010.
  29. Lowea, D., Roya, K., Das, R., Clifton, C.G., Lim, J.B. and Roy, K. (2020), "Full scale experiments on splitting behaviour of concrete slabs in steel concrete composite beams with shear stud connection 2", Struct., 23, 126-138.
  30. Metni, N. and Hamel, T. (2007), "A UAV for bridge inspection: Visual servoing control law with orientation limits", Auto. Constr., 17(1), 3-10. https://doi.org/10.1016/j.autcon.2006.12.010.
  31. Murphy, K.P. (2012), Machine Learning: A Probabilistic Perspective, MIT Press.
  32. Nishikawa, T., Yoshida, J., Sugiyama, T. and Fujino, Y. (2012), "Concrete crack detection by multiple sequential image filtering", Comput.-Aid. Civil Infrastr. Eng., 27(1), 29-47. https://doi.org/10.1111/j.1467-8667.2011.00716.x.
  33. Omar, T. and Nehdi, M.L. (2017), "Remote sensing of concrete bridge decks using unmanned aerial vehicle infrared thermography", Auto. Constr., 83, 360-371. https://doi.org/10.1016/j.autcon.2017.06.024.
  34. Pao, Y. (1989), "Adaptive pattern recognition and neural networks".
  35. Petegrosso, R., Park, S., Hwang, T.H. and Kuang, R. (2017), "Transfer learning across ontologies for phenome-genome association prediction", Bioinform., 33(4), 529-536. https://doi.org/10.1093/bioinformatics/btw649.
  36. Philip, R.E., Andrushia, A.D., Nammalvar, A., Gurupatham, B.G.A. and Roy, K. (2023), "A comparative study on crack detection in concrete walls using transfer learning techniques", J. Compos. Sci., 7(4), 169. https://doi.org/10.3390/jcs7040169.
  37. Rajamony Laila, L., Gurupatham, B.G.A., Roy, K. and BP Lim, J. (2021), "Effect of super absorbent polymer on microstructural and mechanical properties of concrete blends using granite pulver", Struct. Concrete, 22, E898-E915. https://doi.org/10.1002/suco.201900419.
  38. Rajamony Laila, L., Gurupatham, B.G.A., Roy, K. and Lim, J.B. (2021), "Influence of super absorbent polymer on mechanical, rheological, durability, and microstructural properties of self-compacting concrete using non-biodegradable granite pulver", Struct. Concrete, 22, E1093-E1116. https://doi.org/10.1002/suco.201900470.
  39. Ren, C.X., Dai, D.Q., Huang, K.K. and Lai, Z.R. (2014), "Transfer learning of structured representation for face recognition", IEEE Trans. Image Proc., 23(12), 5440-5454. https://doi.org/10.1109/TIP.2014.2365725.
  40. Rimkus, A., Podviezko, A. and Gribniak, V. (2015), "Processing digital images for crack localization in reinforced concrete members", Procedia Eng., 122, 239-243. https://doi.org/10.1016/j.proeng.2015.10.031.
  41. Shin, H.C., Roth, H.R., Gao, M., Lu, L., Xu, Z., Nogues, I., Yao, J., Mollura, D. and Summers, R.M. (2016), "Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics and transfer learning", IEEE Trans. Med. Imag., 35(5), 1285-1298. https://doi.org/10.1109/TMI.2016.2528162.
  42. Simonyan, K. and Zisserman, A. (2014), "Very deep convolutional networks for large-scale image recognition", arXiv preprint arXiv:1409.1556.
  43. Smola, A.J. and Scholkopf, B. (2004), "A tutorial on support vector regression", Statist. Comput., 14, 199-222. https://doi.org/10.1023/B:STCO.0000035301.49549.88.
  44. Sohn, H., Dutta, D., Yang, J., DeSimio, M., Olson, S. and Swenson, E. (2011), "Automated detection of delamination and disbond from wavefield images obtained using a scanning laser vibrometer", Smart Mater. Struct., 20(4), 045017. https://doi.org/10.1088/0964-1726/20/4/045017.
  45. Taha, M.R., Noureldin, A., Lucero, J.L. and Baca, T.J. (2006), "Wavelet transform for structural health monitoring: A compendium of uses and features", Struct. Hlth. Monit., 5(3), 267-295. https://doi.org/10.1177/1475921706067741.
  46. Tajbakhsh, N., Shin, J.Y., Gurudu, S.R., Hurst, R.T., Kendall, C.B., Gotway, M.B. and Liang, J. (2016), "Convolutional neural networks for medical image analysis: Full training or fine tuning?", IEEE Trans. Med. Imag., 35(5), 1299-1312. https://doi.org/10.1109/TMI.2016.2535302.
  47. Tiwari, S., Mondal, G., Dash, S.R. and Roy, K. (2023), "Experimental investigation of unbonded reinforced concrete PT shear wall under lateral loading: A state-of-the-art review", J. Build. Eng., 107504. https://doi.org/10.1016/j.jobe.2023.107504.
  48. Vaghefi, K., Ahlborn, T.M., Harris, D.K. and Brooks, C.N. (2015), "Combined imaging technologies for concrete bridge deck condition assessment", J. Perform. Constr. Facil., 29(4), 04014102. https://doi.org/10.1061/(ASCE)CF.1943-5509.0000465.
  49. Weiss, K., Khoshgoftaar, T.M. and Wang, D. (2016), "A survey of transfer learning", Journal of Big data, 3, 1-40. https://doi.org/10.1186/s40537-016-0043-6.
  50. Wu, X. (2004), "Data mining: artificial intelligence in data analysis", Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology (IAT 2004).
  51. Xu, Q., Xiang, E.W. and Yang, Q. (2010), "Protein-protein interaction prediction via collective matrix factorization", 2010 IEEE International Conference on Bioinformatics and Biomedicine (BIBM).
  52. Zou, Q., Zhang, Z., Li, Q., Qi, X., Wang, Q. and Wang, S. (2018), "Deepcrack: Learning hierarchical convolutional features for crack detection", IEEE Trans. Image Proc., 28(3), 1498-1512. https://doi.org/10.1109/TIP.2018.2878966.