1. Proceedings
  2. I3M
  3. 2022
  4. HMS
  5. 10.46354/i3m.2022.hms.003

Automated damage detection of trailers at intermodal terminals using deep learning 

  • a Pavel Cimili , 
  • b Jana Voegl, 
  • c Patrick Hirsch, 
  • d Manfred Gronalt 
  • a University of Natural Resources and Life Sciences, Vienna, Institute of Production and Logistics, Feistmantelstraße 4, 1180 Vienna, Austria
Cite as
Cimili P., Voegl J., Hirsch P., and Gronalt M. (2022).,Automated damage detection of trailers at intermodal terminals using deep learning. Proceedings of the 24th International Conference on Harbor, Maritime and Multimodal Logistic Modeling & Simulation(HMS 2022). , 003 . DOI: https://doi.org/10.46354/i3m.2022.hms.003

Abstract

Intermodal transport plays a crucial role for sustainable transport in Europe. Inefficiencies at the interface of road and terminal reduce the acceptance of this transport mode. Increasing the efficiency of the gate-in process is therefore vital. An essential part of this process is the damage detection of trailers entering the terminals due to safety and liability issues. It is usually performed by trained personnel. Deep learning promises to assist human resources reliably in this step. While automated damage detection has been discussed in various fields, including container deliveries, trailers stayed out of the research scope. Thus, this work focuses on automatic damage detection of trailers using deep learning. We observe two approaches: transfer and semi-supervised learning. While the first one is based on the MobileNetV2 network, the latter uses convolutional autoencoders and might be helpful not only for detection but also for damage segmentation and visualization. One of the significant difficulties is the trailer detection on images and its partitioning, which is necessary due to the large recording resolution. That is why we also observe a pre-processing algorithm for the real-world images received from an intermodal terminal in Austria.

References

  1. Baldi, P., & Lu, Z. (2012). Complex-valued autoencoders. Neural Networks, 33, 136-147. doi:10.1016/j.neunet.2012.04.011.
  2. Bergmann, P., Löwe, S., Fauser, M., Sattlegger, D., & Steger, C. (2019). Improving unsupervised defect segmentation by applying structural similarity to autoencoders. Proceedings of the 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications 5, 372-380. doi:10.5220/0007364503720380.
  3. Boumessouer, A. (2020, September 20). MVTec-Anomaly-Detection. Retrieved from https://github.com/AdneneBoumessouer/MVTec-Anomaly-Detection.
  4. He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (pp. 770-778). doi:10.1109/CVPR.2016.90.
  5. İmamoğlu, Z. (2019). Container damage detection and classification using container images. Unpublished master's thesis, İzmir Institute of Technology, İzmir, Turkey.
  6. Lin, T. Y., Goyal, P., Girshick, R., He, K., & Dollar, P. (2017). Focal loss for dense object detection. Proceedings of the IEEE international conference on computer vision (pp. 2980-2988). doi:10.1109/ICCV.2017.324.
  7. Mao, X.-J., Shen, C., & Yang, Y.-B. (2016). Image Restoration Using Convolutional Auto-encoders with Symmetric Skip Connections. arXiv preprint arXiv:1606.08921.
  8. Mujeeb, A., Dai, W., Erdt, M., & Sourin, A. (2018). Unsupervised surface defect detection using deep autoencoders and data augmentation. In 2018 International Conference on Cyberworlds (CW) (pp. 391-398). IEEE. doi:10.1109/CW.2018.00076.
  9. Nakazawa, K., Iwasaki, I., & Yamashita, I. (1995). Development of damage detection system for container. Proceedings of IECON'95-21st Annual Conference on IEEE Industrial Electronics (Vol. 2, pp. 1160-1163). IEEE.
  10. Posset, M., Gronalt, M., Peherstorfer, H., Schultze, R.-C., & Starkl, F. (2020). Intermodal Transport Europe. Universität f. Bodenkultur Wien.
  11. Protic, S. M., Fikar, C., Voegl, J., & Gronalt, M. (2020). Analysing the impact of value added services at intermodal inland terminals. International Journal of Logistics Research and Applications, 23(2), 159-177. doi:10.1080/13675567.2019.1657386.
  12. Sarafijanovic-Djukic, N., & Davis, J. (2019). Fast distance-based anomaly detection in images using an inception-like autoencoder. International Conference on Discovery Science (pp. 493-508). Springer, Cham. doi:10.1007/978-3-030-33778-0_37.
  13. Wang, Y., Yoshihashi, R., Kawakami, R., You, S., Harano, T., Ito, M., . . . Naemura, T. (2019). Unsupervised anomaly detection with compact deep features for wind turbine blade images taken by a drone. IPSJ Transactions on Computer Vision and Applications, 11(1). doi:10.1186/s41074-019-0056-0.
  14. Wang, Z., Gao, J., Zeng, Q., & Sun, Y. (2021). Multitype Damage Detection of Container Using CNN Based on Transfer Learning. Mathematical Problems in Engineering, 2021. doi:10.1155/2021/5395494.