Abstract
The main goal of this study is to propose a suitable method in order to reduce the detection cost of orbital defects, improve the efficiency of defect detection, and effectively achieve rail health assessment by combining UAV image acquisition and digital image processing technologies. The characteristics of the images collected by the UAV using the proposed appropriate method to deal with the defects, which Hough transform and horizontal projection method are used to extract the rail area. In the process of image enhancement for rail images, an improved local normalized image enhancement method is proposed, and a defect segmentation method based on maximum entropy threshold value is used for defect segmentation. According to the calculation and extraction of defect characteristics, a classification model is proposed based on the deep forest method to classify the two types of defects, namely spalling and crack. The results show that the proposed method can accurately and effectively classify the rail surface defects in a small sample and has a certain practical reference value.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Shen L (2012) Analysis of the status quo and countermeasures of railway transportation safety. Heilongjiang Sci Technol Inf 22:124 (in Chinese)
Pan G (1910) Discussion on the impact of railway track irregularity on driving safety and stability. Road Bridge Construct (18) (in Chinese)
Liu F (2015) Research on rail irregularity detection and disease identification based on operating vehicles. Beijing Jiaotong University (in Chinese)
Zhang B, Fu X (2001) Types and formation mechanism of tread stripping of railway wheels and wheels. China Railw Sci 02:76–81 (in Chinese)
Yang Jinfeng, Tao Wei (2011) An efficient direction field-based method for the detection of fasteners on high-speed railways. Sensors 11:7364–7381
Utrata D, Clark R (2003) Groundwork for rail flaw detection using ultrasonic phased array inspection. Am Inst Phys, 799–805
Yu H (2013) Research on the detection technology of rail surface defects based on machine vision. Hunan University, Hunan (in Chinese)
Li Q, Ren S (2012) Avisual detection system for rail surface defects. IEEE Trans Syst Man Cybern Part C 42(6):1531–1542
Zhai Y, Yin C, Dang J et al (2016) Rapid identification method of rail region based on image hue value mutation characteristics. J Traffic Transp Eng 16(1) (in Chinese)
Jesussek M, Ellermann K (2015) Fault detection and isolation for a railway vehicle by evaluating estimation residuals. Procedia Iutam 13(1)
Tang X, Wang Y (2013) Vision detection and recognition algorithm for rail surface defects. Comput Eng 39(3) (in Chinese)
Li Q, Ren S (2012) A real time visual inspection system for discrete surface defects of rail heads. IEEE Trans Instrum Meas 61(8)
Li Q, Ren S (2012) A real-time visual inspection system for discrete surface defects of rail heads. IEEE Trans Instrum Meas 61(8):2189–2199
Pun T (1980) A new method for gray-level picture thresholding using the entropy of the histo-gram. Signal Process 2(2):223–237
Pun T (1981) Entropic thresholding: a new approach. Comput Vis Graph Image Process 16(2):210–239
Pun T, Pun T (1980) A new method for gray-level picture threshold using the entropy of the histogram. Sig Process 29(3):273–285
Lu G, Yang K (1996) Morphological characteristics and failure mechanism of rail damage. J China Railw Soc 18(3) (in Chinese)
Liu G, Qiu X (2018) Finger vein recognition based on multi-mode LBP and deep forest. Comput Technol Dev 07:1–6 (in Chinese)
Acknowledgements
This work was supported by the National Key R&D Program of China (No. 2016YFB1200402).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Zhou, Y., Jia, L., Wang, Z., Wu, Y. (2020). Rail Surface Defect Recognition and Classification Method Based on Deep Forest. In: Qin, Y., Jia, L., Liu, B., Liu, Z., Diao, L., An, M. (eds) Proceedings of the 4th International Conference on Electrical and Information Technologies for Rail Transportation (EITRT) 2019. EITRT 2019. Lecture Notes in Electrical Engineering, vol 639. Springer, Singapore. https://doi.org/10.1007/978-981-15-2866-8_48
Download citation
DOI: https://doi.org/10.1007/978-981-15-2866-8_48
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-15-2865-1
Online ISBN: 978-981-15-2866-8
eBook Packages: EngineeringEngineering (R0)