Optics and Precision Engineering, Volume. 33, Issue 3, 438(2025)
Binocular vision-based trackside pantograph anomaly detection under strong environmental noise
Figures & Tables(12)
Fig. 1. Framework of binocular vision-based pantograph abnormality detection system
Fig. 8. Simulation experiment results of laboratory environment noise
Table 1. Algorithm accuracy evaluation in ideal environmental conditions
View tableView in ArticleTable 1. Algorithm accuracy evaluation in ideal environmental conditions
Algorithm Sample Avg ICP RMSE/mm Minimum residual Measured value/mm True value/mm Avg MRE/mm SGM I1/I2/I3/I4 0.24 24.05/24.82/24.78/25.29 23.6/24.6/24.9/25.1 0.25 CostFilter 0.22 23.82/24.47/24.71/25.39 0.21 PatchMatch 0.20 23.44/25.04/25.27/24.82 0.31 DLNR 0.16 23.83/24.18/24.46/24.89 0.28 Selective-IGEV 0.18 23.79/24.81/24.74/24.99 0.17 Ours 0.20 23.36/23.87/24.69/25.22 0.21
Table 2. Accuracy test data of laboratory environment disturbance simulation experiment
View tableView in ArticleTable 2. Accuracy test data of laboratory environment disturbance simulation experiment
Tx/mm Tz/mm Rz/（°） MRE/mm SGM CostFilter PatchMatch DLNR Selective-IGEV Ours 0.9 / / 0.37 0.29 0.36 0.37 -0.16 -0.13 1.8 / / 0.77 0.45 0.69 -0.27 -0.22 -0.20 2.7 / / 1.21 0.99 1.19 0.76 0.31 0.27 3.6 / / 1.53 1.44 1.47 0.83 0.69 0.21 / 0.9 / -0.24 -0.21 -0.27 0.27 -0.22 -0.17 / 1.8 / -0.45 -0.44 -0.55 -0.38 -0.42 0.47 / 2.7 / -0.75 -0.78 -0.69 0.46 -0.40 -0.11 / 3.6 / -1.17 -1.01 -0.99 0.33 -0.68 0.22 / / 0.2 0.45 0.34 0.41 0.43 0.33 0.14 / / 0.4 0.83 0.76 0.75 -0.29 0.59 -0.12 / / 0.6 1.35 1.28 1.20 0.54 0.66 0.16 / / 0.8 1.65 1.54 1.35 0.99 0.64 0.11 1.4 1.4 0.4 1.77 1.52 1.66 0.68 0.43 0.46 / / / 0.56 0.47 0.51 0.36 0.12 0.10 / / / 0.93 0.85 0.85 0.42 0.46 0.22
Table 3. Algorithm accuracy and efficiency evaluation at worksite
View tableView in ArticleTable 3. Algorithm accuracy and efficiency evaluation at worksite
Sample True value/mm MRE/mm SGM CostFilter PatchMatch DLNR Selective-IGEV Ours A 29.40 0.71 0.67 -0.35 0.26 -0.19 -0.12 B 38.90 -0.65 0.79 0.94 0.62 0.43 0.18 C 40.10 0.89 0.74 0.67 0.81 0.63 0.20 D 36.60 -1.31 -0.67 0.95 0.49 -0.33 0.38 E 40.20 0.07 -0.12 0.21 -0.17 0.10 -0.17 F 34.20 -1.14 -0.99 -1.23 -0.46 -0.44 -0.35 G 34.80 -0.41 -0.37 -0.96 -0.37 -0.64 0.34 H 38.20 0.40 0.67 0.37 0.32 0.24 -0.19 Average - 0.70 0.63 0.71 0.44 0.38 0.24 Sample True value/mm ICP RMSE/mm SGM CostFilter PatchMatch DLNR Selective-IGEV Ours A 29.40 1.21 1.77 0.94 0.25 0.13 0.19 B 38.90 0.97 1.45 0.39 0.14 0.27 0.22 C 40.10 1.80 0.39 0.29 0.25 0.47 0.17 D 36.60 0.74 1.88 1.61 0.30 0.65 0.12 E 40.20 0.92 1.32 0.62 0.39 0.24 0.31 F 34.20 1.46 0.34 1.13 0.18 0.04 0.22 G 34.80 1.43 0.03 0.23 0.43 0.08 0.16 H 38.20 1.65 0.95 0.96 0.35 0.62 0.22 Average - 1.27 1.02 0.77 0.29 0.31 0.20 Sample True value/mm Time consume/ms SGM CostFilter PatchMatch DLNR Selective-IGEV Ours A 29.40 112 159 216 2140 2515 79 B 38.90 131 164 222 2266 2529 91 C 40.10 122 163 234 2245 2542 83 D 36.60 127 154 231 2258 2547 90 E 40.20 118 152 214 2226 2498 86 F 34.20 111 155 214 2234 2493 81 G 34.80 124 166 223 2241 2458 86 H 38.20 143 151 233 2247 2504 93 Average - 124 158 223 2232 2511 86
Tools
Get Citation
Copy Citation Text
Jin ZHAO, Yin GUO, Shibin YIN, Lei GUO, Jigui ZHU. Binocular vision-based trackside pantograph anomaly detection under strong environmental noise[J]. Optics and Precision Engineering, 2025, 33(3): 438
Paper Information
Category:
Received: Sep. 2, 2024
Accepted: --
Published Online: Apr. 30, 2025
The Author Email:
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20