13th Asia Pacific Transportation Development Conference
Track Fastener Defect Detection Based on Local Convolutional Neural Networks
Publication: Resilience and Sustainable Transportation Systems
ABSTRACT
Convolutional neural networks (CNN) is one of the typical methods of track fastener state detection. The elastic strip is the main feature used to detect the defects of the track fastener. However, for the detection of the elastic strip, feature extraction is performed on whole pictures through CNN. Many features that are not related to the elastic strip are extracted, which is not conducive to the identification of the state of the elastic strip. Aiming at the above problems, a new method for detecting the track fasteners i.e., local convolutional neural networks (LCNN) is proposed. Firstly, the prior knowledge is used to divide the track fastener picture into two parts: the bolt part and the elastic strip part. The picture of the bolt part is not relevant to the detection, so the feature of the extraction bolt part is not conducive to detection. The main distinguishing feature of the detection is the elastic strip part, so it is more advantageous to use the picture of the elastic strip part to identify the state of the track fastener. Then, the CNN is used to extract the feature of the picture of elastic strip part. The effectiveness of the proposed LCNN for the detection of track fasteners is verified by the image of the track fasteners on the actual railway.
Get full access to this article
View all available purchase options and get full access to this chapter.
ACKNOWLEDGMENTS
The support foundations: National Natural Science Foundation of China(Grant No.51975347), Shanghai Science and Technology Committee(Grant No.18030501300), Shanghai Sailing Program(18YF1409200), Talent Program of Shanghai University of Engineering Science.
REFERENCES
Chen, X., Liu, L., Deng, YB., Kong, XY. (2019). “Vehicle detection based on visual attention mechanism and adaboost cascade classifier in intelligent transportation systems.” Opt Quant Electron., 50(8).
Dou, Y., Huang, Y., and LI, Q. et al. (2014). “A Fast Template Matching-Based Algorithm for Railway Bolts Detection.” Int j Mach Learn Cyb., 5(6),835-844.
Feng, H., Jiang, Z., and Xie, F. et al. (2014). “Automatic Fastener Classification and Defect Detection in Vision -Based Railway Inspection Systems.” IEEE t Instrum Meas., 63(4), 877-888.
Li, Y., Ttinh, H., and Haas, N. et al. (2014). “Rail Component Detection, Optimization, and Assessment for Automatic Rail Track Inspection.” IEEE t Intell Transp., 15(2),760-770.
Liu, J., Wang, K., and Yuan, J. et al. (2014). “Optimization of RBF-SVM Model in Railway Fastener Detection System.” Comput Appl Eng Educ., 50(15), 30-33. In Chinese.
Lecun, Y., Bengio, Y., and Hinton, G. (2015). “Deep learning ”. Nature., 521(7553), 436.
Maeeo, P., Stella, E., and Ancona, N. et al. (2005). “Visual Detection of Hexagonal Headed Bolts Using Method of Frames and Matching Pursuit.” The Second Iberian Conference on Pattern Recognition and Image Analysis., 277-284.
Marino, F., Distance, A., and Mazzeo, P. et al. (2007). “A Real Time Visual Inspection System for Railway Maintenance: Automatic Hexagonal Headed Bolts Detection.” Ieee t Syst Man Cy-s., 37(3), 418-428.
Simard, P., Steinkraus, D., and Platt, J. (2003). “Best practices for Convolutional neural networks applied to visual document analysis.” ICDAR., 958.
Schmidhuber, J. (2014). “Deep learning in neural networks: an overview.” Neural Netw., 61, 85.
Information & Authors
Information
Published In
Resilience and Sustainable Transportation Systems
Pages: 425 - 432
Editors: Fengxiang Qiao, Ph.D., Texas Southern University, Yong Bai, Ph.D., Marquette University, Pei-Sung Lin, Ph.D., University of South Florida, Steven I Jy Chien, Ph.D., New Jersey Institute of Technology, Yongping Zhang, Ph.D., California State Polytechnic University, and Lin Zhu, Ph.D., Shanghai University of Engineering Science
ISBN (Online): 978-0-7844-8290-2
Copyright
© 2020 American Society of Civil Engineers.
History
Published online: Jun 29, 2020
Published in print: Jun 29, 2020
Authors
Metrics & Citations
Metrics
Citations
Download citation
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.