A Computer Vision Based Warning and Behavior Analysis System for Large Passenger Flow in Stations

This paper adopts a microservice architecture to design an intelligent monitoring system for rail transit platforms. The system relies on high-definition video surveillance and utilizes technologies such as video understanding and human posture recognition to comprehensively supervise the entire platform. It uses digital twin technology to model the station and achieve 3D visualization operation and maintenance, displaying real-time passenger flow information and trends, regional passenger flow information and congestion levels, emergency events (fights, falls, lean against platform screen doors, etc.) alarm information and records, and improving passenger safety during boarding.