Research on Steel Rails Defect Detection Method Based on Improved YOLOv5s

WANG Ping

Computer & Telecommunication ›› 2025, Vol. 1 ›› Issue (7) : 13-21.

WANG Ping

Author information +

History +

Abstract

The surface defects of steel rails are difficult to detect due to irregular shapes, scale differences, and background complexity, and the existing YOLOv5s model has limitations such as insufficient bounding box positioning accuracy and weak multi-scale feature extraction ability. This article proposes an improved detection method for YOLOv5s. Firstly, to address the issue of positioning deviation, the GIoU loss function is used to enhance the robustness of bounding box regression; Secondly, embedding CBAM attention mechanism to enhance the focus of defect area features; Finally, to address multi-scale defect detection, the BiFPN structure is introduced to achieve bidirectional weighted fusion. The experiment shows that the improved model can be applied to the dataset of surface defects on steel rails mAP@0.5 reaching 81.8%, an increase of 5.3% compared to the benchmark model, significantly improving the reliability of detection under complex working conditions and providing an effective solution for surface defect detection of rails.

Key words

surface defect detection of steel rails / YOLOv5s / GIoU loss function / CBAM attention mechanism / BiFPN structure

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

WANG Ping. Research on Steel Rails Defect Detection Method Based on Improved YOLOv5s[J]. Computer & Telecommunication. 2025, 1(7): 13-21

References

[1] 罗晖,韩岳霖,马治伟,等.基于双模态融合的钢轨表面缺陷分割研究[J].华东交通大学学报,2025,42(1):52-60.
[2] 周黄健,贾银亮,王平.一种钢轨表面伤损漏磁检测信号滤波方法[J].测控技术,2023,42(5):73-78.
[3] 李军,许炫皓,王耀弘.改进YOLOv5s的钢轨表面缺陷检测算法[J].重庆理工大学学报(自然科学),2024,38(8):130-137.
[4] 白云,谭俊杰,曹林东,等.基于Res Net18-YOLOv8n的地面标志线检测算法[J].电脑与信息技术,2024,32(5):1-6.
[5] 张俊宁,毕泽洋,闫英,等.基于注意力机制与改进YOLO的温室番茄快速识别[J].农业机械学报,2023,54(5):236-243.
[6] 曹风云,李云鹏,包正鹏.融合注意力机制与双向特征加权的PCB缺陷识别[J].武汉理工大学学报,2023,45(6):161-172.
[7] Park J W,Lee T G,Back I C,et al.Rail surface defect detection and analysis using multi-channel eddy current method based algorithm for defect evaluation[J].Journal of Nondestructive Evaluation,2021,40(3):83-94.
[8] 闵永智,岳彪,马宏锋,等.基于图像灰度梯度特征的钢轨表面缺陷检测[J].仪器仪表学报,2018,39(4):220-229.
[9] 陈后金,许文达,郝晓莉.基于灰度-梯度共生矩阵的钢轨表面缺陷检测方法[J].北京交通大学学报,2015,39(2):7-12.
[10] 王国伟,交通信息工程及控制.基于图像处理的钢轨表面缺陷检测方法研究[D].兰州:兰州交通大学,2022.
[11] 韩强,刘俊博,冯其波,等.基于多层级特征融合的钢轨表面伤损检测方法[J].中国铁道科学,2021,42(5):41-49.
[12] 罗晖,徐广隆.基于图像增强与深度学习的钢轨表面缺陷检测[J].铁道科学与工程学报,2021,18(3):623-629.
[13] Zheng J,Han Q,Wang L,et al.Lightweight bimodal rail surface defect detection method:Parallel-YOLOv4-Tiny[C]//Proceedings of the 2021 5th International Conference on Video and Image Processing,2021:98-103.
[14] 陈少璞. 基于机器视觉钢轨表面缺陷检测研究[D].石家庄:石家庄铁道大学,2021.
[15] 崔晓宁,王起才,李盛,等.基于YOLO-v5的双块式轨枕裂缝智能识别[J].铁道学报,2022(4):104-111.
[16] Wei X,Wei D,Suo D,et al.Multi-target defect identification for railway track line based on image processing and improved YOLOv3 model[J].IEEE Access,2020,8:61973-61988.
[17] 何庆,陈正兴,王启航,等.基于改进YOLOV3的钢轨伤损B显图像识别研究[J].铁道学报,2022,44(12):82-88.
[18] Anwar N,Shen Z,Wei Q,et al.YOLOv4 based deep learning algorithm for defects detection and classification of rail surfaces[C]//2021 IEEE International Intelligent Transportation Systems Conference (ITSC).IEEE,2021:1616-1620.
[19] Wang M,Li K,Zhu X,et al.Detection of surface defects on railway tracks based on deep learning[J].IEEE Access,2022,10:126451-126465.
[20] Luo H,Cai L,Li C.Rail surface defect detection based on an improved YOLOv5s[J].Applied Sciences,2023,13(12):7330.
[21] Zhang C,Xu D,Zhang L,et al.Rail surface defect detection based on image enhancement and improved YOLOX[J].Electronics,2023,12(12):2672.
[22] Wang Y,Zhang K,Wang L,et al.An improved YOLOv8 algorithm for rail surface defect detection[J].IEEE Access,2024,12:44984-44997.
[23] Shang L D,Yang Q,Wang J,et al.Detection of rail surface defects based on CNN image recognition and classification[C]//2018 20th International Conference on Advanced Communication Technology (ICACT).IEEE,2018:45-51.
[24] Yu H,Li Q,Tan Y,et al.A coarse-to-fine model for rail surface defect detection[J].IEEE Transactions on Instrumentation and Measurement,2018,68(3):656-666.
[25] 金侠挺. 基于深度学习的高铁钢轨表面缺陷检测方法研究[D].湖南:湖南大学,2020.
[26] 赵宏伟,郑嘉俊,赵鑫欣,等.基于双模态深度学习的钢轨表面缺陷检测方法[J].计算机工程与应用,2023,59(7):285-293.
[27] 孙铁强,魏光辉,宋超,等.基于YOLO的多模态钢轨表面缺陷检测方法[J].电子测量技术,2024,47(21):72-81.
[28] 陈兴. 基于改进YOLOv5的铁轨表面缺陷检测的研究[D].银川:宁夏大学,2024.
[29] 赵春华,罗顺,谭金铃,等.基于PC-YOLOv7算法钢材表面缺陷检测[J].国外电子测量技术,2023,42(9):137-145.