Skip to main content
SpringerLink
Log in

Fault detection of train mechanical parts using multi-mode aggregation feature enhanced convolution neural network

  • Original Article
  • Published:
International Journal of Machine Learning and Cybernetics Aims and scope Submit manuscript

Abstract

Faults in train mechanical parts pose a significant safety hazard to railway transportation. Although some image detection methods have replaced manual fault detection of train mechanical parts, the detection effect on small mechanical parts under low illumination conditions is not ideal. To improve the accuracy and efficiency of the detection of train faults under different environments, we propose a multi-mode aggregation feature enhanced network (MAFENet) based on a single-stage detector (SSD). This network uses the idea of a two-step adjustment structure from coarse to fine and uses the K-means algorithm to design anchors. The receptive field enhancement module (RFEM) is designed to obtain the fusion features of different receptive fields. The attention-guided detail feature enhancement module (ADEM) is designed to complement the detailed features of deep-level feature maps. Meanwhile, the complete intersection over union (CIoU) loss is used to obtain more accurate bounding boxes. The experimental results on the train mechanical parts fault (TMPF) dataset showed that the detection performance of MAFENet is better than those of other SSD models. MAFENet with an input size of 320 × 320 pixels can achieve a mean average precision (mAP) of 0.9787 and a detection speed of 33 frames per second (FPS), which indicates that it can realize real-time detection, has good robustness to images under different environmental conditions, and can be used to improve the efficiency of the detection of faulty train parts.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others

References

  1. Liu R (2005) Principle and application of TFDS. Chinese Railways 5:26–27 ((in Chinese))

    Google Scholar 

  2. Kong R, Sun F, A. Yao, H. Liu, M. Lu, Y. Chen (2017) “Ron: Reverse connection with objectness prior networks for object detection,” in Proc. Comput. Vis. Pattern Recognit. (CVPR), 5936–5944

  3. S. Zhang, L. Wen, X. Bian, Z. Lei, S. Z. Li, (2018) Single-Shot Refinement Neural Network for Object Detection," Proc. IEEE Comput. Vis. Pattern Recognit. (CVPR), pp. 4203–4212

  4. K. He, X. Zhang, S. Ren, J. Sun (2016) Deep Residual Learning for Image Recognition, Proc. IEEE Comput. Vis. Pattern Recognit. (CVPR), pp. 770–778

  5. N. Ma, X. Zhang, H. Zheng, and J. Sun, (2018) ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design, Proc. Springer Euro. Conf. Comput. Vis. (ECCV), pp. 122–138

  6. M. Tan and Q. V. Le (2019) EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks, Proc. Int. Conf. Machine Learning, pp. 6105–6114

  7. Xiao L, Wu B, Hu Y (2020) Surface defect detection using image pyramid. IEEE Sens J 20(13):7181–7188

    Article  Google Scholar 

  8. Zhong J, Liu Z, Han Z, Han Y, Zhang W (2019) A CNN-based defect inspection method for catenary split pins in high-speed railway. IEEE Trans Instrum Meas 68(8):2849–2860

    Article  Google Scholar 

  9. K. He, G. Gkioxari, P. Dollar and R. Girshick (2017) Mask R-CNN," Proc. IEEE Int. Conf. Comput. Vis. (ICCV), pp. 2980–2988

  10. L. Chen, G. Papandreou, F. Schroff, H. J. a. C. V. Adam, and P. Recognition, "Rethinking Atrous Convolution for Semantic Image Segmentation", arXiv: 1706.05587, 2017. [Online]. Available: https://arxiv.org/abs/1706.05587.

  11. S. Mehta, M. Rastegari, A. Caspi, L. G. Shapiro, H. Hajishirzi (2018) ESPNet Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation," Proc. Springer Euro. Conf. Comput. Vis. (ECCV), Springer, Cham 561–580

  12. R. Girshick, J. Donahue, T. Darrell, and J. Malik (2014) Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation Proc. IEEE Comput Vis Pattern Recognit (CVPR), pp. 580–587

  13. S. Ren, K. He, R. Girshick, and J. Sun (2015) Faster R-CNN: towards real-time object detection with region proposal networks, Proc. Neural Inf. Process. Syst. (NIPS), pp. 91–99

  14. T. Lin, P. Dollar, R. Girshick, K. He, B. Hariharan, S. Belongie (2017) Feature Pyramid Networks for Object Detection Proc. IEEE Comput. Vis. Pattern Recognit. (CVPR), pp. 936–944

  15. W. Liu et al., (2016) SSD: Single shot multibox detector, Proc. Springer Euro. Conf. Comput. Vis. (ECCV), pp. 21–37

  16. J. Redmon, S. K. Divvala, R. Girshick, A. Farhadi, (2016) You Only Look Once: Unified, Real-Time Object Detection," Proc. IEEE Comput. Vis. Pattern Recognit. (CVPR), pp. 779–788,

  17. T. Lin, P. Goyal, R. Girshick, K. He, P. Dollar (2017) Focal Loss for Dense Object Detection Proc. Springer Euro. Conf. Comput. Vis. (ECCV), pp. 2999–3007, pp. 318–327

  18. J. Redmon A. Farhadi, "YOLOv3: An incremental improvement", 2018, [online] Available: https://arxiv.org/abs/1804.02767.

  19. Zhou F, Zou R, Qiu Y, Gao H (2014) Automated visual inspection of angle cocks during train operation. Proc Inst Mech Eng Part F-J Rail Rapid Transit 228(7):794–806

    Article  Google Scholar 

  20. Liu L, Zhou F, He Y (2015) Automated status inspection of fastening bolts on freight trains using a machine vision approach. IEEE Trans Instrum Meas 230(7):1629–1641

    Google Scholar 

  21. G. Nan and J. E. Yao (2015) A real-time visual inspection method of fastening bolts in train operation," Proc. Aopc 2015: Image Processing and Analysis 9675

  22. A. James, W. Jie, Y. Xulei, Y. Chenghao, Z. Zeng (2018) TrackNet - A Deep-learning Based Fault Detection for Railway Track Inspection, Proc. 2018 Int. Conf. Intell. Rail Transportation (ICIRT)

  23. R. Singh Pahwa, J. Chao, J. paul (2019) FaultNet: Faulty Rail-Valves Detection using Deep-learning and Computer Vision," Proc. Inst. Mech. Eng. Part F-J. Rail Rapid Transit (IEEE-ITSC), pp. 59–566

  24. Zhou F, Li J, Li X, Li Z, Cao Y (2018) Train target detection in a complex background based on convolutional neural networks. Proc Inst Mech Eng Part F-J Rail Rapid Transit 233(3):298–311

    Article  Google Scholar 

  25. Vaswani A, Shazeer N, Parmar N (2017) Attention is all you need. Proc Neural Inf Process Syst (NIPS) 30:5998–6008

    Google Scholar 

  26. X. Wang, R. Girshick, A. Gupta, K. He (2018) Non-local Neural Networks," Proc. IEEE Comput. Vis. Pattern Recognit. (CVPR), pp. 7794–7803

  27. Yi J, Wu P, Metaxas D, Understanding I (2019) ASSD: Attentive single shot multibox detector. Comput Vis Image Underst 189:1028

    Article  Google Scholar 

  28. Chang C Y, Chang S E, Hsiao P Y, et al. (2020) EPSNet: Efficient Panoptic Segmentation Network with Cross-layer Attention Fusion[C]//Proceedings of the Asian Conference on Computer Vision

  29. J. Yu, Y. Jiang, Z. Wang, Z. Cao, T. S. Huang (2016) UnitBox: An Advanced Object Detection Network," Proc. acm multimedia, pp. 516–520

  30. H. Rezatofighi, N. Tsoi, J. Gwak, A. Sadeghian, I. Reid, S. Savarese (2019) Generalized Intersection Over Union: A Metric and a Loss for Bounding Box Regression," Proc. IEEE Comput. Vis. Pattern Recognit. (CVPR), pp. 658–666

  31. Zheng Z, Wang P, Liu W, Li J, Ye R, Ren D (2020) Distance-IoU Loss: Faster and Better learning for bounding box regression," in national conference on artificial intelligence. Proc AAAI Conf Artif Intell 34:12993–13000

    Google Scholar 

  32. K. Simonyan and A. Zisserman (2014) Very deep convolutional networks for large-scale image recognition,” arXiv: 1409.1556, [Online]. Available: https://arxiv.org/abs/1409.1556.

  33. C. Szegedy et al. (2015) Going deeper with convolutions, Proc. IEEE Comput. Vis. Pattern Recognit. (CVPR), pp. 1–9

  34. S. Liu, D. Huang, Y. Wang (2018) Receptive field block net for accurate and fast object detection, Proc. Springer Euro. Conf. Comput. Vis. (ECCV), Cham 404–419

  35. Hu J, Shen L, Albanie S, Sun G, Wu E (2020) Squeeze-and-excitation networks. IEEE Trans Pattern Anal Mach Intell 42(8):2011–2023

    Article  Google Scholar 

  36. Z. Li, and F. Zhou, "FSSD: Feature Fusion Single Shot Multibox Detector," arXiv: 1712.00960, 2017, [Online]. Available: https://arxiv.org/abs/1712.00960.

  37. Bolei Zhou, Aditya Khosla, Agata Lapedriza Aude Oliva, and Antonio Torralba (2016) “Learning deep features for discriminative localization”. Proc. IEEE CVPR, pp. 2921–2929

Download references

Acknowledgements

We thank China University of Mining and Technology (Beijing) for providing the experimental hardware platform. This work was supported by the National Natural Science Foundation of China (No. 52075027), the Fundamental Research Funds for the Central Universities (2020XJJD03), and the State Key Laboratory of Coal Mining and Clean Utilization, China (2021-CMCU-KF012).

Author information

Authors and Affiliations

  1. China University of Mining and Technology, Bejing, School of Mechanical and Information Engineering, ding 11 Xueyuan Road, Haidian District, Beijing, 100083, China

    Ye Tao, Zhang Jun & Zhang Yi

  2. The State Key Laboratory of Coal Mining and Clean Utilization, Beijing, 100083, China

    Ye Tao & Zhang Jun

  3. The Key Laboratory of Intelligent Mining and Robotics, Ministry of Emergency Management, Beijing, 100083, China

    Ye Tao & Zhang Jun

  4. Unmanned System Department of the 9th Academy of China Aerospace Science and Technology Corporation Limited, China Academy of Aerospace Electronics Technology, Beijing, 100094, China

    Zhang Zhi-hao

  5. School of Instrumentation Science and Opto-Electronics Engineering, Beihang University, Beijing, 100191, China

    Zhou Fu-qiang

  6. Department of Electrical Engineering and Automation, University of Eastern Finland, Kuopio, Finland

    Gao Xiao-zhi

Authors
  1. Ye Tao
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Zhang Jun
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Zhang Zhi-hao
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Zhang Yi
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Zhou Fu-qiang
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Gao Xiao-zhi
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

The overall research objectives and technical routes were formulated by YT. ZJ conducted the investigations, experiments, and wrote the first draft of the manuscript. ZZ-h, ZY, ZF-q, and GX-z assisted in writing the first draft.

Corresponding author

Correspondence to Ye Tao.

Ethics declarations

Conflict of interest

The authors declare that they have no known competing financial interests or personal relationships that could have influenced the work reported in this paper.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Tao, Y., Jun, Z., Zhi-hao, Z. et al. Fault detection of train mechanical parts using multi-mode aggregation feature enhanced convolution neural network. Int. J. Mach. Learn. & Cyber. 13, 1781–1794 (2022). https://doi.org/10.1007/s13042-021-01488-1

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s13042-021-01488-1

Keywords

Search

Navigation