基于机器视觉的均衡地铁客流量研究

Research on Balanced Metro Passenger Flow Based on Machine Vision

地铁车厢乘客拥挤度存在较大差别时,会影响地铁运能的充分发挥。本文以均衡地铁车厢拥挤度为目的,从讨论机器视觉识别车厢拥挤度入手,结合地铁站乘客流量时间分布规律预测地铁各车厢停站实时拥挤度,生成地铁车厢拥挤度分级信息。通过乘客信息系统及时向候车站台乘客发布各节地铁车厢的拥挤度等级信息,引导乘客选择候车位置,使地铁运能能够充分发挥。数据验证表明,高峰期合理的诱导信息,可提高地铁载客能力约10%。

When there is a big difference in the crowding degree of subway passengers,the full play of the subway transportation capacity will be affected.In order to balance the congestion degree of subway cars,this paper starts from discussing machine vision to identify the congestion degree of cars,and predicts the real-time congestion degree of each stop of subway cars in combination with the time distribution law of passenger flow of subway stations,and generates the classification information of the congestion degree of subway cars.The passenger information system promptly releases the information on the congestion level of each subway car to the passengers on the waiting platform,and guides the passengers to select the waiting position,so that the subway transportation can be fully utilized.Data verification shows that reasonable induction information during peak hours can increase subway passenger carrying capacity by about 10%.