摘要
在轨道交通领域客流密度是一项重要监控指标,客流密度评估是客流智能化管控的重要手段,为车站大客流监控提供依据的关键技术。利用车站内传统摄像头,通过图片处理、卷积操作和池化,提出了基于深度学习的视频客流密度计算方法来评估站内区域人数。基于B/S架构设计了客流密度监控平台,通过与实际情况对比,监控平台具有一定的准确性和及时性。
Passenger flow density is an important monitoring indicator in the field of rail transit,and passenger flow density evaluation is an important means of intelligent management of passenger flow,and a key technology that provides a basis for monitoring large passenger flow in stations. The traditional camera is used in the station,through image processing,convolution operation and pooling,a deep learning-based video passenger flow density calculation method is proposed to evaluate the number of people in the station area. Based on the B/S architecture,a passenger flow density monitoring platform is designed. By comparing with the actual situation,the monitoring platform has certain accuracy and timeliness.
作者
黄丰
莫辉强
王伟
欧阳慧
于富洋
张城
叶明
HUANG Feng;MO Huiqiang;WANG Wei;OUYANG Hui;YU Fuyang;ZHANG Cheng;YE Ming(Zhejiang Rail Transit Operation Management Group Co.,Ltd.,Hangzhou 310020;Shenzhen Institute of Beidou Applied Technology,Shenzhen 518055)
出处
《计算机与数字工程》
2022年第10期2149-2152,2165,共5页
Computer & Digital Engineering
关键词
客流密度
深度学习
轨道交通
passenger flow density
deep learning
rail transit
