| Peer-Reviewed

Crack Detection of Concrete Bridges Based on Improved U-Net Model

Received: 2 November 2022     Accepted: 29 November 2022     Published: 8 December 2022
Views:       Downloads:
Abstract

Bridge is an important part of traffic infrastructure, and its maintenance is related to smooth traffic and even the safety of people's lives and property. Automatic detection of concrete bridge cracks is an important part of bridge maintenance. However, it is still a challenging task due to the inhomogeneous strength of concrete bridge cracks, the complexity of the background, and the weakness of the target. Currently, traditional image processing-based technology cannot detect bridge cracks well, and the typical deep learning model is not very effective when directly used to detect bridge cracks. In response to this problem, this paper proposes a bridge crack detection model based on the improved U-Net deep learning model. Compared with the original U-Net model, the convolution block of our improved model is wider in the encoding stage, and the wider convolution block enables the model to learn more semantically discriminative information. Finally, the performance of the improved model is validated on a dataset with 600 images. Experimental results show that compared with other advanced detection models, the proposed method achieves better performance in terms of average intersection ratio and pixel accuracy. Therefore, the method proposed in this paper can detect concrete bridge cracks automatically, which provides a certain reference value for bridge crack maintenance engineers based on image recognition technology.

Published in Science Discovery (Volume 10, Issue 6)
DOI 10.11648/j.sd.20221006.29
Page(s) 500-505
Creative Commons

This is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium or format, provided the original work is properly cited.

Copyright

Copyright © The Author(s), 2022. Published by Science Publishing Group

Keywords

Deep Learning, Bridge Crack Detection, U-Net, Convolutional Neural Networks

References
[1] ZHANG Lei, YANG Fan, ZHANG Daniel-Yimin, et al. Road Crack Detection Using Deep Convolutional Neural Network [C]//Proceedings of IEEE International Conference on Image Processing. Phoenix, AZ, USA: IEEE, 2016: 3708-3712.
[2] JTG 5120-2021, 公路桥涵养护规范 [S].
[3] ABDEL-QADER Ikhlas, PASHAIE-RAD Sara, ABUDAYYEH Osama, et al. PCA-based Algorithm for Unsupervised Bridge Cack Detection [J]. Advances in Engineering Software, 2006, 37 (12): 771-778.
[4] YEUM Chul-Min, DYKE Shirley J. Vision-based Automated Crack Detection for Bridge Inspection [J]. Computer-Aided Civil and Infrastructure Engineering, 2015, 30 (10): 759-770.
[5] PRASANNA Prateek, DANA Kristin-J., GUCUNSKI Nenad, et al. Automated Crack Detection on Concrete Bridges [J]. IEEE Transactions on Automation Science and Engineering, 2014, 13 (2): 591-599.
[6] CHA Young-Jin, CHOI Wooram, BUYUKOZTURK Oral. Deep Learning-based Crack Damage Detection Using Convolutional Neural Networks [J]. Computer-Aided Civil and Infrastructure Engineering, 2017, 32 (5): 361-378.
[7] 张维峰,刘萌,杨明慧.基于数字图像处理的桥梁裂缝检测技术[J].现代交通技术, 2008, 5 (5): 34-36.
[8] 胡世昆. 基于数字图像处理技术的路面裂缝检测算法研究[D]. 南京:南京邮电大学, 2012.
[9] KRIZHEVSKY Alex, SUTSKEVER Ilya, HINTON Geoffrey E. ImageNet Classification with Deep Convolutional Neural Networks [J]. Communications of the ACM, 2012, 60 (6): 84-90.
[10] SZEGEDY Christian, LIU Wei, JIA Yang-qing, et al. Going Deeper with Convolutions [C]// Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Boston, MA, USA: IEEE, 2015: 1-9.
[11] REN Shao-qing, HE Kai-ming, GIRSHICK Ross, et al. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39 (6): 1137-1149.
[12] RONNEBERGER Olaf, FISCHER Phillip, BROX Thomas. U-Net: Convolutional Networks for Biomedical Image Segmentation [C]//NAVAB N, HORNEGGER J, WELLS W M, et al. Lecture Notes in Computer Science: Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015. Cham: Springer, 2015: 234-241.
[13] ZAGORUYKO Sergey, KOMODAKIS Nikos. Wide Residual Networks [C]//Proceedings of the British Machine Vision Conference. York: BMVA Press, 2016: 87.1-87.12.
[14] 李良福, 马卫飞, 李丽, 等. 基于深度学习的桥梁裂缝检测算法研究. 自动化学报, 2019, 45 (9): 1727-1742.
[15] PASZKE Adam, GROSS Sam, MASSA Francisco, et al. PyTorch: An Imperative Style, High-Performance Deep Learning Library [C]//Proceedings of the 33rd International Conference on Neural Information Processing Systems. NY, USA: Red Hook, 2019: 8026–8037.
[16] CHEN Liang-Chieh, PAPANDREOU George, KOKKINOS Iasonas, et al. DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40 (4): 834-848.
[17] LONG Jonathan, SHELHAMER Evan, DARRELL Trevor. Fully Convolutional Networks for Semantic Segmentation [C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Boston, USA: IEEE, 2015: 3431-3440.
[18] ZHAO Heng-shuang, SHI Jian-ping, QI Xiao-juan, et al. Pyramid Scene Parsing Network [C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, USA: IEEE, 2017: 6230-6239.
Cite This Article
  • APA Style

    Zhen Huang, Xingtuo Zhang, Kaizhong Xie, Xiaojun Ke, Yiyi Zhang. (2022). Crack Detection of Concrete Bridges Based on Improved U-Net Model. Science Discovery, 10(6), 500-505. https://doi.org/10.11648/j.sd.20221006.29

    Copy | Download

    ACS Style

    Zhen Huang; Xingtuo Zhang; Kaizhong Xie; Xiaojun Ke; Yiyi Zhang. Crack Detection of Concrete Bridges Based on Improved U-Net Model. Sci. Discov. 2022, 10(6), 500-505. doi: 10.11648/j.sd.20221006.29

    Copy | Download

    AMA Style

    Zhen Huang, Xingtuo Zhang, Kaizhong Xie, Xiaojun Ke, Yiyi Zhang. Crack Detection of Concrete Bridges Based on Improved U-Net Model. Sci Discov. 2022;10(6):500-505. doi: 10.11648/j.sd.20221006.29

    Copy | Download

  • @article{10.11648/j.sd.20221006.29,
      author = {Zhen Huang and Xingtuo Zhang and Kaizhong Xie and Xiaojun Ke and Yiyi Zhang},
      title = {Crack Detection of Concrete Bridges Based on Improved U-Net Model},
      journal = {Science Discovery},
      volume = {10},
      number = {6},
      pages = {500-505},
      doi = {10.11648/j.sd.20221006.29},
      url = {https://doi.org/10.11648/j.sd.20221006.29},
      eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.sd.20221006.29},
      abstract = {Bridge is an important part of traffic infrastructure, and its maintenance is related to smooth traffic and even the safety of people's lives and property. Automatic detection of concrete bridge cracks is an important part of bridge maintenance. However, it is still a challenging task due to the inhomogeneous strength of concrete bridge cracks, the complexity of the background, and the weakness of the target. Currently, traditional image processing-based technology cannot detect bridge cracks well, and the typical deep learning model is not very effective when directly used to detect bridge cracks. In response to this problem, this paper proposes a bridge crack detection model based on the improved U-Net deep learning model. Compared with the original U-Net model, the convolution block of our improved model is wider in the encoding stage, and the wider convolution block enables the model to learn more semantically discriminative information. Finally, the performance of the improved model is validated on a dataset with 600 images. Experimental results show that compared with other advanced detection models, the proposed method achieves better performance in terms of average intersection ratio and pixel accuracy. Therefore, the method proposed in this paper can detect concrete bridge cracks automatically, which provides a certain reference value for bridge crack maintenance engineers based on image recognition technology.},
     year = {2022}
    }
    

    Copy | Download

  • TY  - JOUR
    T1  - Crack Detection of Concrete Bridges Based on Improved U-Net Model
    AU  - Zhen Huang
    AU  - Xingtuo Zhang
    AU  - Kaizhong Xie
    AU  - Xiaojun Ke
    AU  - Yiyi Zhang
    Y1  - 2022/12/08
    PY  - 2022
    N1  - https://doi.org/10.11648/j.sd.20221006.29
    DO  - 10.11648/j.sd.20221006.29
    T2  - Science Discovery
    JF  - Science Discovery
    JO  - Science Discovery
    SP  - 500
    EP  - 505
    PB  - Science Publishing Group
    SN  - 2331-0650
    UR  - https://doi.org/10.11648/j.sd.20221006.29
    AB  - Bridge is an important part of traffic infrastructure, and its maintenance is related to smooth traffic and even the safety of people's lives and property. Automatic detection of concrete bridge cracks is an important part of bridge maintenance. However, it is still a challenging task due to the inhomogeneous strength of concrete bridge cracks, the complexity of the background, and the weakness of the target. Currently, traditional image processing-based technology cannot detect bridge cracks well, and the typical deep learning model is not very effective when directly used to detect bridge cracks. In response to this problem, this paper proposes a bridge crack detection model based on the improved U-Net deep learning model. Compared with the original U-Net model, the convolution block of our improved model is wider in the encoding stage, and the wider convolution block enables the model to learn more semantically discriminative information. Finally, the performance of the improved model is validated on a dataset with 600 images. Experimental results show that compared with other advanced detection models, the proposed method achieves better performance in terms of average intersection ratio and pixel accuracy. Therefore, the method proposed in this paper can detect concrete bridge cracks automatically, which provides a certain reference value for bridge crack maintenance engineers based on image recognition technology.
    VL  - 10
    IS  - 6
    ER  - 

    Copy | Download

Author Information
  • Key Laboratory of Disaster Prevention and Structural Safety of Ministry of Education, Guangxi University, Nanning, China

  • Key Laboratory of Disaster Prevention and Structural Safety of Ministry of Education, Guangxi University, Nanning, China

  • Key Laboratory of Disaster Prevention and Structural Safety of Ministry of Education, Guangxi University, Nanning, China

  • Key Laboratory of Disaster Prevention and Structural Safety of Ministry of Education, Guangxi University, Nanning, China

  • Key Laboratory of Disaster Prevention and Structural Safety of Ministry of Education, Guangxi University, Nanning, China

  • Sections