Rail transit clearing method based on deep learning

Abstract

The invention provides a rail transit clearing method based on deep learning, and the method comprises the steps: firstly obtaining the video information of a passenger entering a station, extracting the optimal face frame and the optimal human body frame of the passenger according to the video information, binding the optimal face frame and the optimal human body frame, obtaining the face and human body features of the passenger, reducing a search database, and improving the recognition accuracy. And finally, comprehensively judging according to the running time data of the rail transit train to accurately obtain the travel track of the passenger. By means of the method, the accuracy of the travel track of the passenger can be effectively improved, meanwhile, the query request can be initiated at any time, and the timeliness is higher.

Description

technical field [0001] The present disclosure relates to the technical field of rail traffic monitoring, in particular to a deep learning-based rail traffic sorting method. Background technique [0002] In the current urban rail transit, the requirements for networked operation of urban rail transit, the need for the convenience of passengers to travel and transfer, and the needs for income distribution of different operating entities determine that in the mode of barrier-free transfer, the urban rail transit clearing and sorting center (ACC) should provide accurate and timely ticket clearing services for all urban rail transit lines and operators. [0003] Passenger OD route: The OD route in the rail transit industry means that passengers start from the station A, go through N (N=0, 1, 2,...) transfers, pass through N transfer stations, and finally exit from station B Outbound, all transfer routes. [0004] The existing sorting system: through the arrival timetable of all...

AI Technical Summary

Problems solved by technology

[0005] The current methods for predicting passenger OD paths mainly have the following problems: 1. There is a strong correlation with the passenger’s arrival timetable. When the timetable deviates, the prediction results will change accordingly; 2. The real-time performance is not enough, and it is impossible to know a passenger in real time OD path, which station it currently appears in; 3. It has a strong correlation with the walking speed of passengers in the transfer station. When passengers stay, go to the bathroom, etc., there will be errors when predicting which train the passenger may catch , which in turn causes the passenger OD path prediction accuracy to decline

Images