Abstract: The high pedestrian flow and complex environment in the entering / exiting scenarios of railway passenger stations make existing tail detection methods have a high false alarm rate. For this reason, this paper designed a tracking detection system for railway passenger entering and exiting station. The paper used structured light 3D ranging technology, identity recognition technology and pedestrian detection technology to accurately obtain the relative distance between passengers. Compared with existing tailgate detection systems, the designed system can effectively detect the tailgate behavior of passenger entering / exiting of station scenarios, significantly improving the detection rate of tailgate situations and reducing the tailgate false alarm rate when passengers carry large luggage.