Railway Track Detection Based on Vehicle Laser Point Cloud

Weigang Li; Yang Mei; Xiang Fan; Yuntao Zhao

doi:10.3788/CJL202249.0410002

Chinese Journal of Lasers, Volume. 49, Issue 4, 0410002(2022)

Railway Track Detection Based on Vehicle Laser Point Cloud

Weigang Li^*, Yang Mei, Xiang Fan, and Yuntao Zhao

Engineering Research Center of Metallurgical Automation and Measurement Technology, Ministry of Education, Wuhan University of Science and Technology, Wuhan, Hubei 430081, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(12)

Fig. 1. Flow of track detection algorithm

Download full size

Fig. 2. Flow of Euclidean clustering based on elevation constraints

Download full size

Fig. 3. Rail track diagram. (a) Track section; (b) rail section

Download full size

Fig. 4. Sleeper details. (a) Top view of non-bridge area; (b) side view of non-bridge area; (c) top view of bridge area; (d) side view of bridge area

Download full size

Fig. 5. Extraction effect of subgrade area

Download full size

Fig. 6. Detection effect of rail surface point cloud under different grid sizes. (a)(f) Grid size is 0.04 m; (b)(g) grid size is 0.06 m; (c)(h) grid size is 0.08 m; (d)(i) grid size is 0.10 m; (e)(j) grid size is 0.12 m

Download full size

Fig. 7. Effect of sleeper point cloud detection in different areas. (a) Non-bridge area; (b) bridge area

Download full size

Fig. 8. Point cloud detection effect of rail surface and sleeper in different areas. (a)(d) Effect of rail surface extraction; (b)(e) effect of sleeper extraction; (c)(f) overall effect

Download full size

Table 1. Accuracy of rail extraction under different grid sizes　unit: %

View table

Table 1. Accuracy of rail extraction under different grid sizes　unit: %

Evaluation indicator	d_grid＝0.04 m		d_grid＝0.06 m		d_grid＝0.08 m		d_grid＝0.10 m		d_grid＝0.12 m
Evaluation indicator	Area 1	Area 2	Area 1	Area 2	Area 1	Area 2	Area 1	Area 2	Area 1	Area 2
r	75.6	50.1	96.2	96.0	98.0	99.2	98.2	99.4	98.2	61.1
p	98.2	99.8	98.6	99.3	99.5	99.3	99.6	99.4	99.7	99.7
q	75.5	50.0	95.0	95.0	97.5	98.4	97.9	98.7	98.0	61.0

Table 2. Comparison of extraction results of rail surface using two methods　unit: %
View table
Table 2. Comparison of extraction results of rail surface using two methods　unit: %
Evaluation indicator Yang’s method Proposed method
Area 1 Area 2 Average Area 1 Area 2 Average
r 95.0 99.1 97.1 98.5 99.5 99.0
p 99.4 98.9 99.2 98.0 99.4 99.1
q 94.5 98.1 96.3 96.6 98.9 97.8

Table 3. Accuracy of sleeper extraction in different basic thresholds d_th of non-bridge area　unit: %
View table
Table 3. Accuracy of sleeper extraction in different basic thresholds d_th of non-bridge area　unit: %
Evaluation indicator d_th＝0.25 m d_th＝0.26 m d_th＝0.27 m d_th＝0.28 m d_th＝0.29 m d_th＝0.30 m d_th＝0.31 m
r 4.9 39.5 88.6 99.4 99.8 99.8 99.9
p 99.9 99.5 98.9 93.6 88.1 61.1 48.2
q 4.9 39.5 87.8 93.1 87.8 61.1 48.2

Table 4. Accuracy of sleeper extraction in different basic thresholds d_th of bridge area　unit: %
View table
Table 4. Accuracy of sleeper extraction in different basic thresholds d_th of bridge area　unit: %
Evaluation indicator d_th＝0.22 m d_th＝0.23 m d_th＝0.24 m d_th＝0.25 m d_th＝0.26 m d_th＝0.27 m d_th＝0.28 m
r 48.2 78.6 96.9 99.3 99.5 99.7 99.9
p 93.9 94.8 95.8 94.7 91.8 88.3 80.9
q 46.8 75.3 92.9 94.1 91.4 88.1 80.8

Tools

Get Citation

Copy Citation Text

Weigang Li, Yang Mei, Xiang Fan, Yuntao Zhao. Railway Track Detection Based on Vehicle Laser Point Cloud[J]. Chinese Journal of Lasers, 2022, 49(4): 0410002

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Received: Jun. 1, 2021

Accepted: Jul. 30, 2021

Published Online: Jan. 18, 2022

The Author Email: Li Weigang (liweigang.luck@foxmail.com)

DOI:10.3788/CJL202249.0410002

Topics

laser devices and laser physics

Lasers and Laser Optics

Laser physics

laser manufacturing

Instrumentation, Measurement and Metrology

Table 1. Accuracy of rail extraction under different grid sizes unit: %

Table 1. Accuracy of rail extraction under different grid sizes unit: %

Table 2. Comparison of extraction results of rail surface using two methods unit: %

Table 2. Comparison of extraction results of rail surface using two methods unit: %

Table 3. Accuracy of sleeper extraction in different basic thresholds dth of non-bridge area unit: %

Table 3. Accuracy of sleeper extraction in different basic thresholds dth of non-bridge area unit: %

Table 4. Accuracy of sleeper extraction in different basic thresholds dth of bridge area unit: %

Table 4. Accuracy of sleeper extraction in different basic thresholds dth of bridge area unit: %

Table 1. Accuracy of rail extraction under different grid sizes　unit: %

Table 1. Accuracy of rail extraction under different grid sizes　unit: %

Table 2. Comparison of extraction results of rail surface using two methods　unit: %

Table 2. Comparison of extraction results of rail surface using two methods　unit: %

Table 3. Accuracy of sleeper extraction in different basic thresholds d_th of non-bridge area　unit: %

Table 3. Accuracy of sleeper extraction in different basic thresholds d_th of non-bridge area　unit: %

Table 4. Accuracy of sleeper extraction in different basic thresholds d_th of bridge area　unit: %

Table 4. Accuracy of sleeper extraction in different basic thresholds d_th of bridge area　unit: %