Abstract：Due to human factors or natural disasters, the phenomenon of foreign matter invading urban rail transit tunnels sometimes occurs. If the driver's emergency response is not sensitive, serious accidents will occur. Therefore, it is important to monitor the operating environment in real time to realize the identification of foreign objects and train control. In order to intelligently invade the detection system and improve the operational reliability, this paper is based on the image acquisition module and the ranging module, uses the improved background frame difference method to apply the image detection method when the foreign object invades, and discusses the active safety behavior of the train. At the same time, the Matlab simulation is carried out by using the video data provided by Nanjing Metro, and the laboratory verification is carried out to confirm the feasibility of this method. Moreover, the foreign object detecting device has low hardware cost and high detection precision, and can provide a basis for active safety behavior of urban rail transit.
2019, Vol. 20 
