Rapid detection of wheel tread defects for YOLO-v5 trains based on residual attention

ZHANG Changfan; XU Yifu; HE Jing; YANG Haonan

doi:10.13890/j.issn.1000-128X.2022.06.001

您当前的位置：

首页 >

文章列表页 >

Rapid detection of wheel tread defects for YOLO-v5 trains based on residual attention

Railway Rolling Stock | 更新时间：2024-08-02

- Rapid detection of wheel tread defects for YOLO-v5 trains based on residual attention
- Electric drive for locomotives Issue 6, Pages: 1-9(2022)
- 作者机构：
  
  湖南工业大学电气与信息工程学院，湖南株洲　412007
- 作者简介：
- 基金信息：
- DOI：10.13890/j.issn.1000-128X.2022.06.001
  CLC： U270.331⁺.1
- Published：10 November 2022，
  
  Received：28 May 2022，
- 稿件说明：
扫描看全文
ZHANG Changfan, XU Yifu, HE Jing, et al. Rapid detection of wheel tread defects for YOLO-v5 trains based on residual attention. [J]. Electric drive for locomotives (6):1-9(2022)
DOI：

ZHANG Changfan, XU Yifu, HE Jing, et al. Rapid detection of wheel tread defects for YOLO-v5 trains based on residual attention. [J]. Electric drive for locomotives (6):1-9(2022) DOI： 10.13890/j.issn.1000-128X.2022.06.001.

摘要

为实现快速准确地检测轮对踏面缺陷，针对轮对踏面噪声干扰大、传统检测算法特征融合不充分的问题，提出一种基于残差注意力的YOLO-v5列车轮对踏面缺陷快速检测方法。首先，针对噪声干扰大的问题，设计了一个残差注意力降噪模块，以有效提升模型检测准确率，并使用Grad-CAM类激活映射技术验证残差注意力模块降低噪声干扰的作用；其次，针对特征融合不充分和模型容易产生漏检的问题，使用一种双向特征金字塔特征融合模块，对主干网络提取的特征进行高效融合，从而有效地降低检测漏检率；最后，采集了数百幅轮对踏面真实缺陷图像，并与5种经典检测模型进行对比，验证了算法的优越性。试验结果表明，该算法能够达到77.9%的准确率和72.3%的召回率，同时所提算法的图像检测速度能达到125幅/秒，模型权重仅为15.1 MB。该模型能快速准确地检测出剥离和凹陷2种缺陷，可便捷地应用于实际的轮对踏面实时缺陷检测场景。

Abstract

In respect of large noise interference of wheel tread and insufficient feature fusion of traditional detection algorithms

in order to achieve fast and accurate detection of wheel tread defects

a method for rapid detection of wheel tread defects for YOLO-v5 trains based on residual attention was proposed. First

for the large noise interference

a residual attention noise reduction module was designed to effectively improve the accuracy of model detection

and the Grad-CAM activation mapping technology was used to verify the effect of the residual attention module on reducing noise interference. Secondly

in view of insufficient feature fusion and the model being prone to omission

a feature fusion module of bidirectional feature pyramid was used to efficiently fuse the features extracted by the backbone network

thereby effectively reducing the omission rate. Finally

hundreds of real defect images of wheel treads were collected and contrasted with five classic detection models to verify the superiority of the algorithm. The results show that the algorithm can achieve 77.9% accuracy and 72.3% recall; in addition

the proposed algorithm can detect 125 images per second

and the model weight is only 15.1 MB. The model can quickly and accurately detect such two defects as peeling and sagging

and can be easily applied to the actual real-time defect detection scenario of wheel tread.

关键词

缺陷检测轮对踏面YOLO-v5残差注意力双向特征金字塔

Keywords

defect detectionwheel treadYOLO-v5residual attentionbidirectional feature pyramid

references

张向阳, 罗江平. 钢轨探伤车综合智能检测系统[J]. 机车电传动, 2021(1): 133-137.

ZHANG Xiangyang, LUO Jiangping. Integrated intelligent system for rail flaw detection vehicle[J]. Electric Drive for Locomotives, 2021(1): 133-137.

罗江平, 喻熙倬, 曹经纬, 等. 基于深度学习与支持向量机的钢轨伤损智能识别系统[J]. 机车电传动, 2021(2): 100-107.

LUO Jiangping, YU Xizhuo, CAO Jingwei, et al. Intelligent rail flaw detection system based on deep learning and support vector machine[J]. Electric Drive for Locomotives, 2021(2): 100-107.

王雪, 付振波. 采用小波分析与支持向量机的车轮踏面擦伤识别方法[J]. 中国机械工程, 2004, 15(18): 1641-1643.

WANG Xue, FU Zhenbo. A new method of wavelet and support vector machine for detection of the train wheel bruise[J]. China Mechanical Engineering, 2004, 15(18): 1641-1643.

徐志军, 陈建政. 基于图像处理和Hough变换的轮对踏面轮廓检测方法[J]. 电子测量技术, 2017, 40(6): 117-121.

XU Zhijun, CHEN Jianzheng. Tread profile of wheel detection method based image processing and Hough transform[J]. Electronic Measurement Technology, 2017, 40(6): 117-121.

SHENG Zhenwen, WANG Guiyun. Fast method of detecting packaging bottle defects based on ECA-EfficientDet[J]. Journal of Sensors, 2022, 2022: 9518910.

REN Shaoqing, HE Kaiming, GIRSHICK R, 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.

REDMON J, FARHADI A. Yolov3: an incremental improvement[DB/OL]. (2018-04-08) [2022-05-12]. https://arxiv.org/abs/1804.02767https://arxiv.org/abs/1804.02767.

BOCHKOVSKIY A, WANG C W, LIAO H Y M. YOLOv4: optimal speed and accuracy of object detection[DB/OL]. (2020-04-23) [2022-05-12]. https://arxiv.org/abs/2004.10934https://arxiv.org/abs/2004.10934.

顾超越, 李喆, 史晋涛, 等. 基于改进Faster-RCNN的无人机巡检架空线路销钉缺陷检测[J]. 高电压技术, 2020, 46(9): 3089-3096.

GU Chaoyue, LI Zhe, SHI Jintao, et al. Detection for pin defects of overhead lines by UAV patrol image based on improved Faster-RCNN[J]. High Voltage Engineering, 2020, 46(9): 3089-3096.

ZHANG Changfan, HU Xinliang, HE Jing, et al. Yolov4 high-speed train wheelset tread defect detection system based on multiscale feature fusion[J]. Journal of Advanced Transportation, 2022, 2022: 1172654.

郝帅, 杨磊, 马旭, 等. 基于注意力机制与跨尺度特征融合的YOLOv5输电线路故障检测[J/OL]. 中国电机工程学报: 1-12. (2022-01-28) [2022-05-06]. http://kns.cnki.net/kcms/detail/11.2107.tm.20220126.1718.008.htmlhttp://kns.cnki.net/kcms/detail/11.2107.tm.20220126.1718.008.html.

HAO Shuai, YANG Lei, MA Xu, et al. YOLOv5 transmission line fault detection based on attention mechanism and cross-scale feature fusion[J/OL]. Proceedings of the CSEE: 1-12. (2022-01-28) [2022-05-06]. http://kns.cnki.net/kcms/detail/11.2107.tm.20220126.1718.008.htmlhttp://kns.cnki.net/kcms/detail/11.2107.tm.20220126.1718.008.html.

WANG C Y, LIAO H Y M, WU Y H, et al. CSPNet: a new backbone that can enhance learning capability of CNN[C]//IEEE. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Seattle, WA: IEEE, 2020: 1571-1580.

RAMACHANDRAN P, ZOPH B, LE Q V. Searching for activation functions[DB/OL]. (2017-10-27) [2022-05-24]. https://arxiv.org/abs/1710.05941https://arxiv.org/abs/1710.05941.

LIN T Y, DOLLÁR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//IEEE. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, HI: IEEE, 2017: 936-944.

LIU Shu, QI Lu, QIN Haifang, et al. Path aggregation network for instance segmentation[C]//IEEE. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 8759-8768.

刘迪, 贾金露, 赵玉卿, 等. 基于深度学习的图像去噪方法研究综述[J]. 计算机工程与应用, 2021, 57(7): 1-13.

LIU Di, JIA Jinlu, ZHAO Yuqing, et al. Overview of image denoising methods based on deep learning[J]. Computer Engineering and Applications, 2021, 57(7): 1-13.

HU Jie, SHEN Li, SUN Gang. Squeeze-and-excitation networks[C]//IEEE. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 7132-7141.

WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//Springer. Computer Vision – ECCV 2018. Munich: Springer, 2018: 3-19.

ZHU Ke, WU Jianxin. Residual attention: a simple but effective method for multi-label recognition[C]//IEEE. 2021 IEEE/CVF International Conference on Computer Vision (ICCV). Montreal: IEEE, 2021: 184-193.

TAN Mingxing, PANG Ruoming, LE Q V. EfficientDet: scalable and efficient object detection[C]//IEEE. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle: IEEE, 2020: 10778-10787.

金楷荣, 王俊平, 陈胜蓝. 基于YOLOv3-tiny的列车轮对踏面缺陷定位与测量方法[J]. 控制与信息技术, 2022(2): 69-75.

JIN K R, WANG J P, CHEN S L. A method of locating and measuring train wheel tread defects based on YOLOv3-tiny[J]. Control and Information Technology, 2022(2): 69-75.

李淼成, 王俊平, 沈云波, 等. 基于迭代重加权最小二乘直线拟合的车轮踏面动态检测基准定位方法[J]. 控制与信息技术, 2022(1): 89-96.

LI M C, WANG J P, SHEN Y B, et al. Wheel tread dynamic detection benchmark positioning method based on iterative reweighted least-squares line fitting[J]. Control and Information Technology, 2022(1): 89-96.

Views

下载量

CSCD

CNKI被引量

Alert me when the article has been cited

提交

Tools

Publicity Resources

No data

Related Author

No data

Related Institution

No data

⁰