Abstract
Autonomous driving train is an important component in the future rail transit system, as it can greatly improve the efficiency and safety of train operation. The most critical part in self-driving train is the active perception, due to the fact there’s no active sensing system in the existing train control systems today, we hereby develop a “Train Intelligent Detection System” (“TIDS”) to achieve reliable and real-time capable environment perception. The TIDS system consists three modules to simulate the real-world transit environment. The first module is to recognize the rail area by using semantic segmentation. In the second module, a convolutional neural network (“CNN”) is used to identify the train in the image. We then use the third module to judge if the identified train affects normal train operation on the recognized rail area. The detection result of TIDS system has been tested in multiple real-world scenarios including but not limited to tunnel and turnouts environment, the results so far have been stable and positive.
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
Selver, M.A, Er, E.: Camera based driver support system for rail extraction using 2-D Gabor wavelet decompositions and morphological analysis. In: Proceedings of ICIRT, Birmingham, UK, pp. 270–275 (2016)
Wang, Z., Cai, B., Jia, C.: Geometry constraints-based visual rail track extraction. In: Proceedings of WCICA, Guilin, China, pp. 993–998 (2016)
Krizhevsky. A., Sutskever, I., Hinton, G.: ImageNet classification with deep convolutional neural networks. In: Proceedings of NIPS, Nevada, USA, pp. 1097–1105 (2017)
Szegedy. C., Liu, W., Jia, Y.: Going deeper with convolutions. In: Proceedings of CVPR, Boston, MA, pp. 1–9 (2015)
Nassu, B.T., Ukai, M.: A Vision-based approach for rail extraction and its application in a camera pan-tilt control system. IEEE Trans. Intell. Transp. Syst. 13(4), 1763–1771 (2012)
Ruder, M, Mohler, N., Ahmed, F.: An obstacle detection system for automated trains. In: Proceedings of IEEE IV, Columbus, OH, USA, pp. 180–185 (2003)
Selver, M.A., Zoral, E.Y.: Predictive modeling for monocular vision based rail track extraction. In: Proceedings of IEEE CISP-BMEI 2017, Shanghai, China, pp. 1–6 (2017)
Wang, Z., Wu, X., Yu, G.: Efficient rail area detection using convolutional neural network. IEEE Access 6, 77656–77664 (2018)
Li, X., Liu, Z., Luo, P.: Not all pixels are equal: difficulty-aware semantic segmentation via deep layer cascade. In: Proceedings of CVPR, Honolulu, HI, USA, pp. 6459–6468 (2017)
Ren, S., He, K., Ross, G.: Faster R-CNN: towards real-time object detection with region proposal networks. IEEE Trans. Pattern Anal. Mach. Intell. 39(6), 1137–1149 (2017)
He, K., Zhang, X., Ren, S.: Deep residual learning for image recognition. In: Proceedings of CVPR, Las Vegas, NV, USA, pp. 770–778 (2016)
Hu, C., Wang, Y., Yu, G., Wang, Z.: Embedding CNN-based fast obstacles detection for autonomous vehicles. In: Proceedings of Intelligent and Connected Vehicles Symposium, Kunshan, China, pp. 1–11 (2018)
Liu, W., Anguelov, D., Erhan, D.: SSD: single shot multibox detector. In: Proceedings of ECCV, Amsterdam, The Netherlands, pp. 21–37 (2016)
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Yang, Z., Cheung, V., Gao, C., Zhang, Q. (2020). Train Intelligent Detection System Based on Convolutional Neural Network. In: Cassenti, D. (eds) Advances in Human Factors and Simulation. AHFE 2019. Advances in Intelligent Systems and Computing, vol 958. Springer, Cham. https://doi.org/10.1007/978-3-030-20148-7_15
Download citation
DOI: https://doi.org/10.1007/978-3-030-20148-7_15
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-20147-0
Online ISBN: 978-3-030-20148-7
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)