The monitoring of subway entry speed is mainly done by subway hardware equipment such as transponders in the existing subway control system. In order to establish external feedback that does not rely on subway hardware equipment, a visual measurement algorithm of subway entry speed based on light reflection is proposed based on the video of entering subway on the platform. After converting the subway entry video into an image sequence, firstly use the reflection phenomenon of the fixed light source on the subway compartment to segment each frame of image. Then, according to the different changes of the reflective pixel area in the door and window of the subway subway, the door frame and the window frame are detected in the image sequence. Finally, the length of the carriage and the number of frames that the carriage passes through the fixed light source are used to determine the subway's entry speed. This method is used to conduct experiments for the network acquisition of video. Through analysis and calculation, the average speed of the subway when the first to sixth carriages pass the light reflection position are 36.6m/s, 33m/s, 31.4m/s, 26.4m/ s, 20.6m/s, 9.6m/s. This shows that this method can realize the detection of subway entry speed, thereby establishing closed-loop feedback outside the subway hardware.

Ruiqi Liu, Duozhi Qin, Bo Hu. Speed Measurement Algorithm for Subway Entering Station Based on Light Reflection. International Journal of Frontiers in Engineering Technology (2020), Vol. 2, Issue 3: 55-62. https://doi.org/10.25236/IJFET.2020.020306.

[1] Kouroussis Georges, Kinet Damien, Moeyaert Veronique, Dupuy Julien and Caucheter Christophe (2016). Rail-way structure monitoring solutions using fibre Bragg grating sensors. International Journal of Rail Transportation, vol.4, no.3, p. 135-150.
[2] Wang Qiang (2019). Application of auxiliary speed measurement and positioning technology of LKJ-15 Train operation monitoring system. Electric Drive For Locomotives,  no.1, p.126-128+133.
[3] Tong Xinglin, Wei Jingchuang, Deng Chengwei, Zhang Cui, Mao Yan, Xu Ouyang and Yin Renjie (2014). Research progress on speed measurement and positioning technology of high-speed maglev train. Laser Journal, vol.41, no.3, p.1-5.
[4] Bjorn Lidestam, Lars Eriksson and Olle Eriksson (2019). Speed perception affected by field of view: Energy-based versus rhythm-based processing. Transportation Research Part F: Traffic Psychology and Behaviour, vol.65, no.8, p.227-241.
[5] Saleh Javadi, Mattias Dahl and Mats I (2019). Vehicle speed measurement model for video-based systems. Computers & Electrical Engineering, vol.76, no.6, p.238-248.
[6] Gustavo A.P. de Morais, Lucas B. Marcos, Joes Nuno A.D. Bueno, Nilo F. de Resende, Marco Henrique Terra and Valdir Grassi Jr (2020). Vision-based robust control framework based on deep reinforcement learning applied to autonomous ground vehicles. Control Engineering Practice, vol.104, no.4, p.1-14.
[7] Yu Zhou and Ronggang Cao (2019). The artificial neural network prediction algorithm research of rail-gun current and armature speed based on B-dot probes array. Measurement, vol.133, no.2, p.47-55.
[8] Hou Wenjun, Huang Wei and Fumamulin (2019). A study of subway vehicle image detection system based on pattern recognition technology. Railway Quality Control, no.5, p.44-47+51.
[9] Sun Pei and Hu Xiaomin (2020). Metro passenger flow statistics system based on image feature extraction. Machinery & Electronics,  no.10, p.17-20.
[10] Zhou Lixue and Ma Chengqian (2019). Real time detection and simulation of urban road tunnel object. Computer Simulation, no.1, p.192-196.
[11] Zhao Wengeng, Zhang Xu and Shao Qingqing (2020). Automatic point cloud registration method of subway tunnels based on reflection intensity images. Bulletin of Surveying and Mapping, no.9, p.38-41.