Rail Surface Defect Detection Based on An Improved YOLOv5s
Source abstract
Abstract
As the operational time of the railway increases, rail surfaces undergo irreversible defects. Once the defects occur, it is easy for them to develop rapidly, which seriously threatens the safe operation of trains. Therefore, the accurate and rapid detection of rail surface defects is very important. However, in the detection of rail surface defects, there are problems, such as low contrast between defects and the background, large scale differences, and insufficient training samples. Therefore, we propose a rail surface defect detection method based on an improved YOLOv5s in this paper. Firstly, the sample dataset of rail surface defect images was augmented with flip transformations, random cropping, and brightness transformations. Next, a Conv2D and Dilated Convolution(CDConv) module was designed to reduce the amount of network computation. In addition, the Swin Transformer was combined with the Backbone and Neck ends to improve the C3 module of the original network. Then, the global attention mechanism (GAM) was introduced into PANet to form a new prediction head, namely Swin transformer and GAM Prediction Head (SGPH). Finally, we used the Soft-SIoUNMS loss to replace the original CIoU loss, which accelerates the convergence speed of the algorithm and reduces regression errors. The experimental results show that the improved YOLOv5s detection algorithm reaches 96.9% in the average precision of rail surface defect detection, offering the accurate and rapid detection of rail surface defects, which has certain engineering application value.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Infrastructure Maintenance and Monitoring · Railway Engineering and Dynamics · Industrial Vision Systems and Defect Detection
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Information Engineering, East China Jiaotong University, Nanchang 330013, China
Funding
National Natural Science Foundation of China
References (39)
Show source-provided references
1. 3D characterization of rolling contact fatigue crack networks10.1016/j.wear.2016.06.027 ↗
2. Detection and evaluation of rail surface defects using alternating current field measurement techniques10.1177/0954409712444840 ↗
3. Jiang, Y., Wang, H., Tian, G., Chen, S., Zhao, J., Liu, Q., and Hu, P. (2018, January 6–8). Non-contact ultrasonic detection of rail surface defects in different depths. Proceedings of the 2018 IEEE Far East NDT New Technology & Application Forum (FENDT), Xiamen, China.10.1109/FENDT.2018.8681948 ↗
4. Use of a 3D model to improve the performance of laser-based railway track inspection10.1177/0954409718795714 ↗
5. Velocity effect analysis of dynamic magnetization in high speed magnetic flux leakage inspection10.1016/j.ndteint.2014.02.001 ↗
6. Zhang, Y., Luo, L., Zhang, Y., Gao, X., and Long, J. (2019, January 6–9). Interlaced scanning by laser ultrasonic for defects imaging of train rail surface. Proceedings of the Eleventh International Conference on Information Optics and Photonics (CIOP 2019), Xi’an, China.10.1117/12.2547455 ↗
7. A review of research on detection and evaluation of the rail surface defects10.12700/APH.19.3.2022.3.14 ↗
8. Maximally Stable Extremal Region Marking-Based Railway Track Surface Defect Sensing10.1109/JSEN.2016.2615333 ↗
9. Rail image segmentation based on Otsu threshold method10.3788/OPE.20162407.1772 ↗
10. A Real-Time Visual Inspection System for Discrete Surface Defects of Rail Heads10.1109/TIM.2012.2184959 ↗
11. Research on inverse P-M diffusion-based rail surface defect detection
12. Background differencing-based high-speed rail surface defect image segmentation
13. Improved Sobel algorithm for defect detection of rail surfaces with enhanced efficiency and accuracy10.1007/s11771-016-3350-3 ↗
14. Vision saliency detection of rail surface defects based on PCA model and color features
15. Research on Rail Surface Defect Detection Method Based on Gray Equalization Model Combined with Gabor Filter
16. Girshick, R., Donahue, J., Darrell, T., and Malik, J. (2014, January 23–28). Rich feature hierarchies for accurate object detection and semantic segmentation. Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition, Columbus, OH, USA.10.1109/CVPR.2014.81 ↗
17. Fast R-CNN
18. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks10.1109/TPAMI.2016.2577031 ↗
19. Faghih-Roohi, S., Hajizadeh, S., Nunez, A., Babuska, R., and De Schutter, B. (2016, January 24–29). Deep Convolutional Neural Networks for Detection of Rail Surface Defects. Proceedings of the International Joint Conference on Neural Networks (IJCNN), Vancouver, BC, Canada.10.1109/IJCNN.2016.7727522 ↗
20. Deep Multitask Learning for Railway Track Inspection10.1109/TITS.2016.2568758 ↗
21. Shang, L., Yang, Q., Wang, J., Li, S., and Lei, W. (2018, January 11–14). Detection of rail surface defects based on CNN image recognition and classification. Proceedings of the 2018 20th International Conference on Advanced Communication Technology (ICACT), Chuncheon, Republic of Korea.10.23919/ICACT.2018.8323642 ↗
22. Yu, C., Deng, H.G., and Feng, Y.X. (2020, January 12–14). Effects of Faster Region-based Convolutional Neural Network on the Detection Efficiency of Rail Defects under Machine Vision. Proceedings of the 2020 IEEE 5th Information Technology and Mechatronics Engineering Conference (ITOEC), Chongqing, China.
23. DM-RIS: Deep Multi model Rail Inspection System with Improved MRF-GMM and CNN10.1109/TIM.2019.2909940 ↗
24. Redmon, J., Divvala, S., Girshick, R., and Farhadi, A. (2016, January 27–30). You Only Look Once: Unified, Real-Time Object Detection. Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA.10.1109/CVPR.2016.91 ↗
25. Redmon, J., and Farhadi, A. (2017, January 21–26). YOLO9000: Better, Faster, Stronger. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA.10.1109/CVPR.2017.690 ↗
26. Redmon, J., and Farhadi, A. (2018, January 18–23). YOLOv3: An incremental improvement. Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Salt Lake City, UT, USA.
27. Bochkovskiy, A., Wang, C.Y., and Liao, H.Y. (2020). YOLOv4: Optimal Speed and Accuracy of Object Detection. arXiv.
28. Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C.Y., and Berg, A.C. (2016). SSD: Single Shot Multi Box Detector. European Conference on Computer Vision, Springer.10.1007/978-3-319-46448-0_2 ↗
29. Song, Y.N., Zhang, G.H., Liu, L., and Zhang, H. (December, January 30). Rail surface defect detection method based on YOLOv3 deep learning networks. Proceedings of the CAC 2018: Proceedings of the 2018 Chinese Automation Congress, Xi’an, China.
30. Multi-Target Defect Identification for Railway Track Line Based on Image Processing and Improved YOLOv3 Model10.1109/ACCESS.2020.2984264 ↗
31. Zhao, J., Zhang, X., Yan, J., Qiu, X., Yao, X., Tian, Y., Zhu, Y., and Cao, W. (2021). A Wheat Spike Detection Method in UAV Images Based on Improved YOLOv5. Remote Sens., 13.10.3390/rs13163095 ↗
32. Zhu, X., Lyu, S., Wang, X., and Zhao, Q. (2021, January 11–17). TPH-YOLOv5: Improved YOLOv5 Based on Transformer Prediction Head for Object Detection on Drone-captured Scenarios. Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW), Montreal, BC, Canada.10.1109/ICCVW54120.2021.00312 ↗
33. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, Ł., and Polosukhin, I. (2017, January 4–9). Attention Is All You Need. Proceedings of the 31st International Conference on Neural Information Processing Systems (NIPS’17), Long Beach, CA, USA.
34. Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., and Guo, B. (2021, January 11–17). Swin Transformer: Hierarchical vision Transformer using shifted windows. Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision (ICCV), Montreal, BC, Canada.10.1109/ICCV48922.2021.00986 ↗
35. Global Attention Mechanism: Retain Information to Enhance Channel Spatial Interactions
36. Woo, S., Park, J., Lee, J.Y., and Kweon, I.S. (2018, January 8–14). CBAM: Convolutional block attention module. Proceedings of the European Conference on Computer Vision, Munich, Germany.10.1007/978-3-030-01234-2_1 ↗
37. Gevorgyan, Z. (2022). SIoU loss: More powerful learning for bounding box regression. arXiv.
38. Bodla, N., Singh, B., Chellappa, R., and Davis, L.S. (2017, January 22–29). Soft-NMS—Improving Object Detection With One Line of Code. Proceedings of the 2017 IEEE International Conference on Computer Vision ICCV, Venice, Italy.10.1109/ICCV.2017.593 ↗
39. Yolox: Exceeding yolo series in 2021