Research on deep learning method for rail surface defect detection

被引:49
作者
Feng, Jiang Hua [1 ]
Yuan, Hao [1 ]
Hu, Yun Qing [1 ]
Lin, Jun [1 ]
Liu, Shi Wang [1 ]
Luo, Xiao [1 ]
机构
[1] CRRC Zhuzhou Inst Co Ltd, Shidai Rd, Zhuzhou, Peoples R China
关键词
rails; feature extraction; object detection; learning (artificial intelligence); computer vision; neural nets; railway engineering; novel object detection algorithm; rail defects; novel detection layers; defects detection; deep learning method; rail surface defect detection; rail transportation system; traditional machine vision algorithms; complex rail surface defects; classification methods; complex deep convolutional networks; defects localisation; real-time processing; VISUAL INSPECTION SYSTEM;
D O I
10.1049/iet-est.2020.0041
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
Rail surface defect detection plays a critical role in the maintenance of the rail transportation system. Video analysis technology is a promising method to detect defects due to its low cost and effectiveness. Several attempts with hand-craft features have been made to obtain the detection results by using traditional machine vision algorithms. However, these methods suffer from imprecise results due to challenging conditions, such as deteriorated and changeable lighting environment and various types of complex rail surface defects. Recently, classification methods with complex deep convolutional networks have become popular. Despite their high accuracy, these methods cannot meet the requirements of defects localisation and real-time processing in practice. To solve these problems, this study proposes a novel object detection algorithm to detect rail defects. The net architecture of the proposed algorithm includes a backbone network using MobileNet and several novel detection layers with multi-scale feature maps inspired by you only look once (YOLO) and feature pyramid networks. Two different architectures of MobileNet are used to estimate the performance of defects detection. The experimental results demonstrate the great potential of the proposed algorithm with fast inference speed and high accuracy in the industry.
引用
收藏
页码:436 / 442
页数:7
相关论文
共 50 条
[31]   Defect detection: Defect Classification and Localization for Additive Manufacturing using Deep Learning Method [J].
Han, Feng ;
Liu, Sheng ;
Liu, Sheng ;
Zou, Jingling ;
Ai, Yuan ;
Xu, Chunlin .
2020 21ST INTERNATIONAL CONFERENCE ON ELECTRONIC PACKAGING TECHNOLOGY (ICEPT), 2020,
[32]   Green Plums Surface Defect Detection Based on Deep Learning Methods [J].
Zhou, Chenxin ;
Wang, Honghong ;
Liu, Yang ;
Ni, Xiaoyu ;
Liu, Ying .
IEEE ACCESS, 2022, 10 :100397-100407
[33]   Research on Defect Detection of Railway Key Components Based on Deep Learning [J].
Zhao B. ;
Dai M. ;
Li P. ;
Ma X. ;
Wu Y. .
Tiedao Xuebao/Journal of the China Railway Society, 2019, 41 (08) :67-73
[34]   Rail defect detection using ultrasonic surface waves [J].
Edwards, RS ;
Jian, X ;
Fan, Y ;
Dixon, S .
REVIEW OF PROGRESS IN QUANTITATIVE NONDESTRUCTIVE EVALUATION, VOLS 25A AND 25B, 2006, 820 :1601-1608
[35]   A Deep Extractor for Visual Rail Surface Inspection [J].
Zhang, Ziwen ;
Liang, Mangui ;
Wang, Zhe .
IEEE ACCESS, 2021, 9 :21798-21809
[36]   FS-RSDD: Few-Shot Rail Surface Defect Detection with Prototype Learning [J].
Min, Yongzhi ;
Wang, Ziwei ;
Liu, Yang ;
Wang, Zheng .
SENSORS, 2023, 23 (18)
[37]   Image-Based Surface Defect Detection Using Deep Learning: A Review [J].
Bhatt, Prahar M. ;
Malhan, Rishi K. ;
Rajendran, Pradeep ;
Shah, Brual C. ;
Thakar, Shantanu ;
Yoon, Yeo Jung ;
Gupta, Satyandra K. .
JOURNAL OF COMPUTING AND INFORMATION SCIENCE IN ENGINEERING, 2021, 21 (04)
[38]   Faster Metallic Surface Defect Detection Using Deep Learning with Channel Shuffling [J].
Yasir, Siddiqui Muhammad ;
Ahn, Hyunsik .
CMC-COMPUTERS MATERIALS & CONTINUA, 2023, 75 (01) :1847-1861
[39]   DSANet-KD: Dual Semantic Approximation Network via Knowledge Distillation for Rail Surface Defect Detection [J].
Zhou, Wujie ;
Hong, Jiankang ;
Ran, Xiaoxiao ;
Yan, Weiqing ;
Jiang, Qiuping .
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, 2024, 25 (10) :13849-13862
[40]   RSDNet: A New Multiscale Rail Surface Defect Detection Model [J].
Du, Jingyi ;
Zhang, Ruibo ;
Gao, Rui ;
Nan, Lei ;
Bao, Yifan .
SENSORS, 2024, 24 (11)