浏览全部资源
扫码关注微信
纸质出版日期： 2021-02-16 ，
录用日期： 2020-06-11
移动端阅览
王建柱, 李清勇, 张靖, 甘津瑞. 轨道病害视觉检测：背景、方法与趋势[J]. 中国图象图形学报, 2021,26(2):287-296.
Jianzhu Wang, Qingyong Li, Jing Zhang, Jinrui Gan. Visual inspection of rail defects: background, methodologies, and trends[J]. Journal of Image and Graphics, 2021,26(2):287-296.
铁路作为国家重要基础设施、国民经济大动脉和大众化运输方式，对社会经济发展起着不可替代的支撑作用。轨道是铁路系统的重要组件，轨道病害检测是铁路工务部门的核心业务。传统的人工巡检不仅费时费力，而且检测结果容易受到各种主观因素的影响。因此，自动化轨道病害检测对维护铁路运输安全具有重要的现实意义。考虑到视觉检测在速度、成本和可视化等方面的优势，本文聚焦于轨道病害视觉检测技术。首先以广泛应用的无砟轨道为例介绍轨道的基本结构，对常见的轨道表观病害进行样例展示、成因分析和影响评价；简要梳理常见的自动化轨道检测技术的基本原理和应用场景，对轨道病害视觉检测面临的图像质量不均、可用特征较少和模型更新困难等主要挑战进行归纳；然后，依照前景模型、背景模型、盲源分离模型及深度学习模型的分类逻辑对轨道病害视觉检测领域的研究现状进行综述，简要介绍了各类方法的代表性工作，总结了各类方法的技术特点与应用局限性；最后，针对智能化铁路的发展需求，展望了未来轨道病害视觉检测技术的研究趋势，即利用小样本/零样本学习、多任务学习与多源异构数据融合等技术手段来解决当前视觉检测系统中的鲁棒性弱、虚警率高等问题。
As a national critical infrastructure
economic artery
and popular transportation
a railway plays an irreplaceable role in supporting the economic and social development of a nation. The rail is the key component of a railway
and correspondingly
rail defect detection is a core activity in railway engineering. Traditional manual inspection is time-consuming and laborious
and its results are easily influenced by various subjective factors. Therefore
automatic defect inspection for maintaining railway safety is highly significant. Considering the advantages of visual inspection in terms of speed
cost
and visualization
this study focuses on machine vision-based techniques. The track structure is first introduced by using the widely used ballastless track as an example. Sample presentation
causal analysis
and impact assessment of typical surface defects are provided. Then
the basic principles and application scenarios of common automatic rail defect detection technologies are briefly reviewed. In particular
ultrasonic techniques can be used to detect rail internal flaws
but it can hardly inspect fatigue damage on rail surface because of factors
such as ultrasonic reflection. Furthermore
detection speed is typically unsatisfactory. Eddy current inspection can obtain information about rail surface defects with the use of a detection coil by measuring the variance of eddy currents generated by an excitation coil. In contrast with ultrasonic technology
eddy current testing is fast and exhibits a distinct advantage in terms of detecting defects
such as shelling and scratch. However
it fails at finding defects that are located at the rail waist and base. Consequently
eddy current detection is frequently used in conjunction with ultrasonic equipment. Notably
eddy current inspection has high requirements for the installation position of the detection coil and the actual operation. Debugging the equipment is a complicated task
and the stability of the detection result is insufficient. Thereafter
current major challenges in the visual inspection of rail defects
namely
inhomogeneity of image qualities
limitation of available features
and difficulty in model updating
are summarized. Then
research actuality in the visual inspection of rail defects is systematically reviewed by categorizing the techniques into foreground
background
blind source separation
and deep learning-based models. One or two representative studies are elaborated in each category
followed by the analysis of technical features and practical limitations. In particular
foreground models typically suppress disturbing noise through operations
such as local image filtering
which can enhance contrast between the defect and the background
and thus
help recognize rail surface defects. This type of models generally exhibits low computational complexity
and thus
can meet the requirements of real-time inspection. However
they easily generate false positives and can hardly segment the defect target. Instead of directly placing emphasis on the defect
background methods model the image background by utilizing the spatial consistency and continuity of the rail image. Similar to foreground models
such methods also exhibit good real-time performance
but effectively decreasing false detection still requires further research. Blind source separation models detect rail defects on the basis of the low rank of the image background and the sparseness of the defect. Compared with the aforementioned two types of models
these approaches do not simply rely on the low-level visual characteristics of the defect target. However
these models tend to require high computational complexity. Deep learning-based models generally exhibit promising performance in the visual inspection of rail defects. However
training a deep learning model frequently requires a large amount of samples
and collecting and labeling numerous defect images can be costly. Moreover
these approaches typically depend on a dataset with specific supervision information
and thus
they may not perform well in other similar scenarios. Finally
future research trends in the visual inspection of rail defects are prospected by targeting the development requirements of smart railways. That is
technologies
such as few-shot or zero-shot learning
multitask learning
and multisource heterogeneous data fusion
should be explored to solve the problems of weak robustness and high false alarm rate existing in current visual inspection systems.
轨道病害视觉检测前景背景盲源分离深度学习
rail defectsvisual inspectionforegroundbackgroundblind source separationdeep learning
Dai P, Wang S C and Gu Z C. 2019. Intelligent image analysis technology and application for rail track inspection//Zhou Y L, Wahab M A, Maia N M M, Liu L Y and Figueiredo E, eds. Data Mining in Structural Dynamic Analysis. Singapore: Springer: 85-99[DOI: 10.1007/978-981-15-0501-0_5http://dx.doi.org/10.1007/978-981-15-0501-0_5]
Dong B Y, Li Q Y, Wang J Z, Huang W, Dai P and Wang S C. 2019. An end-to-end abnormal fastener detection method based on data synthesis//Proceedings of the 31st IEEE International Conference on Tools with Artificial Intelligence. Portland, USA: IEEE: 149-156[DOI: 10.1109/ICTAI.2019.00029http://dx.doi.org/10.1109/ICTAI.2019.00029]
Faghih-Roohi S, Hajizadeh S, Núñez A, Babuska R and De Schutter B. 2016. Deep convolutional neural networks for detection of rail surface defects//Proceedings of 2016 International Joint Conference on Neural Networks. Vancouver, Canada: IEEE: 2584-2589[DOI: 10.1109/ijcnn.2016.7727522http://dx.doi.org/10.1109/ijcnn.2016.7727522]
Gan J R, Li Q Y, Wang J Z and Yu H M. 2017. A hierarchical extractor-based visual rail surface inspection system. IEEE Sensors Journal, 17(23): 7935-7944[DOI: 10.1109/JSEN.2017.2761858]
Gan J R. 2019. Visual Inspection Models and Algorithms for Rail Surface Defects. Beijing: Beijing Jiaotong University
甘津瑞. 2019.轨道表面缺陷的视觉检测模型与算法研究.北京: 北京交通大学
Gan J R, Wang J Z, Yu H M, Li Q Y and Shi Z P. 2020. Online rail surface inspection utilizing spatial consistency and continuity. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 50(7): 2741-2751[DOI: 10.1109/TSMC.2018.2827937]
Gibert X, Patel V M and Chellappa R. 2017. Deep multitask learning for railway track inspection. IEEE Transactions on Intelligent Transportation Systems, 18(1): 153-164[DOI: 10.1109/TITS.2016.2568758]
He Z D, Wang Y N, Liu J and Yin F. 2016. Background differencing-based high-speed rail surface defect image segmentation. Chinese Journal of Scientific Instrument, 37(6): 640-649
贺振东, 王耀南, 刘洁, 印峰. 2016.基于背景差分的高铁钢轨表面缺陷图像分割.仪器仪表学报, 37(6): 640-649)[DOI: 10.19650/j.cnki.cjsi.2016.03.022]
He Z D, Wang Y N, Mao J X and Yin F. 2014. Research on inverse P-M diffusion-based rail surface defect detection. Acta Automatica Sinica, 40(8): 1667-1679
贺振东, 王耀南, 毛建旭, 印峰. 2014.基于反向P-M扩散的钢轨表面缺陷视觉检测.自动化学报, 40(8): 1667-1679)[DOI: 10.3724/SP.J.1004.2014.01667]
Jin X T, Wang Y N, Zhang H, Zhong H, Liu L, Wu Q M J and Yang Y M. 2020. DM-RIS: deep multimodel rail Inspection system with improved MRF-GMM and CNN. IEEE Transactions on Instrumentation and Measurement, 69(4): 1051-1065[DOI: 10.1109/TIM.2019.2909940]
Li B X, Dai X X, Wang K C P, Yang E H and Peng Y. 2016. Automated survey of railway conditions: a preliminary investigation//Proceedings of International Conference on Transportation and Development. Houston, USA: ASCE: 446-454[DOI: 10.1061/9780784479926.041http://dx.doi.org/10.1061/9780784479926.041]
Li Q Y, Liang Z P, Huang Y P and Shi Z Z. 2014. Sparseness representation model for defect detection and its application. Journal of Computer Research and Development, 51(9): 1929-1935
李清勇, 梁正平, 黄雅平, 史忠植. 2014.缺陷检测的稀疏表示模型及应用.计算机研究与发展, 51(9): 1929-1935)[DOI: 10.7544/issn1000-1239.2014.20140153]
Li Q Y and Ren S W. 2012a. A real-time visual inspection system for discrete surface defects of rail heads. IEEE Transactions on Instrumentation and Measurement, 61(8): 2189-2199[DOI: 10.1109/TIM.2012.2184959]
Li Q Y and Ren S W. 2012b. A visual detection system for rail surface defects. IEEE Transactions onSystems, Man, and Cybernetics, Part C (Applications and Reviews), 42(6): 1531-1542[DOI: 10.1109/TSMCC.2012.2198814]
Li Q Y, Zhong Z D, Liang Z P and Liang Y. 2015. Rail inspection meets big data: methods and trends//Proceedings of the 18th International Conference on Network-Based Information Systems. Taipei, China: IEEE: 302-308[DOI: 10.1109/NBiS.2015.47http://dx.doi.org/10.1109/NBiS.2015.47]
Liu J B, Huang Y P, Zou Q, Tian M, Wang S C, Zhao X X, Dai P and Ren S W. 2019a. Learning visual similarity for inspecting defective railway fasteners. IEEE Sensors Journal, 19(16): 6844-6857[DOI: 10.1109/JSEN.2019.2911015]
Liu J J, Li B L, Xiong Y, He B and Li L. 2015. Integrating the symmetry image and improved sparse representation for railway fastener classification and defect recognition. Mathematical Problems in Engineering, 2015: #462528[DOI: 10.1155/2015/462528]
Liu S, Wang Q D and Luo Y P. 2019b. A review of applications of visual inspection technology based on image processing in the railway industry. Transportation Safety and Environment, 1(3): 185-204[DOI: 10.1093/tse/tdz007]
Liu Z, Wang W and Wang P. 2010. Design of machine vision system for inspection of rail surface defects. Journal of Electronic Measurement and Instrument, 24(11): 1012-1017
刘泽, 王嵬, 王平. 2010.钢轨表面缺陷检测机器视觉系统的设计.电子测量与仪器学, 24(11): 1012-1017)[DOI: 10.3724/SP.J.1187.2010.01012]
Loveday P W, Long C S and Ramatlo D A. 2019. Ultrasonic guided wave monitoring of an operational rail track. Structural Health Monitoring, 19(16): #147592171989358[DOI: 10.1177/1475921719893887]
Min Y Z, Xiao B Y, Dang J W, Yue B and Cheng T D. 2018. Real time detection system for rail surface defects based on machine vision. Eurasip Journal on Image and Video Processing, 2018(1): #3[DOI: 10.1186/s13640-017-0241-y]
Papaelias M P, Roberts C and Davis C. 2008. A review on non-destructive evaluation of rails: state-of-the-art and future development. Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and rapid transit, 222(4): 367-384[DOI: 10.1243/09544097 JRRT209]
Romera-Paredes B and Torr P. 2015. An embarrassingly simple approach to zero-shot learning//Proceedings of the 32nd International Conference on Machine Learning. Lille, France: ICML: 2152-2161
Snell J, Swersky K and Zemel R. 2017. Prototypical networks for few-shot learning//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach, USA: ACM: 4080-4090
Thomas H M, Heckel T and Hanspach G. 2007. Advantage of a combined ultrasonic and eddy current examination for railway inspection trains. Insight-Non-Destructive Testing and Condition Monitoring, 49(6): 341-344[DOI: 10.1784/insi.2007.49.6.341]
Wang J Z, Li Q Y, Gan J R, Yu H M and Yang X. 2020. Surface defect detection via entity sparsity pursuit with intrinsic priors. IEEE Transactions on Industrial Informatics, 16(1): 141-150[DOI: 10.1109/TII.2019.2917522]
Wang X M, Ni W B and Wang P. 2013. Overview and future development of rails nondestructive inspection. Nondestructive Testing, 35(2): 10-17
王雪梅, 倪文波, 王平. 2013.高速铁路轨道无损探伤技术的研究现状和发展趋势.无损检测, 35(2): 10-17
Yu H M, Li Q Y, Tan Y Q, Gan J R, Wang J Z,Geng Y A and Jia L. 2019. A coarse-to-fine model for rail surface defect detection. IEEE Transactions on Instrumentation and Measurement, 68(3): 656-666[DOI: 10.1109/TIM.2018.2853958]
Zhang L N and Cen Y G. 2019. Low-rank matrix decomposition for rail track image defect detection. Journal of Signal Processing, 35(4): 667-675
张琳娜, 岑翼刚. 2019.铁轨图像的低秩矩阵分解缺陷检测.信号处理, 35(2): 667-675)[DOI: 10.16798/j.issn.1003-0530.2019.04.018]
Zhang Yand Yang Q. 2017. A survey on multi-task learning[EB/OL].[2020-04-01].https:arxiv.org/pdf/1707.08114.pdfhttps:arxiv.org/pdf/1707.08114.pdf
Zhang Y G, Ning G Q and Mao S H. 2011. Analysis and monitoring of railway flaking defects. Railway Quality Control, 39(10): 30-31
张永革, 宁广庆, 毛胜辉. 2011.钢轨鱼鳞伤分析及探测监控.铁道技术监督, 39(10): 30-31)[DOI: 10.3969/j.issn.1006-9178.2011.10.014]
Zhou J H, Ren A C, Ji Y, Zhang Y H and Liu F S. 2013. Analysis of the reasons for the spalling defects on U71Mn rail treads. China Railway Science, 34(2): 1-6
周剑华, 任安超, 吉玉, 张银花, 刘丰收. 2013. U71Mn钢轨踏面剥离掉块缺陷分析.中国铁道科学, 34(2): 1-6)[DOI: 10.3969/j.issn.1001-4632.2013.02.01]
相关作者
相关机构
京公网安备11010802024621