基于视频识别的智能交通信号灯配时优化  

Intelligent Traffic Signal Timing Optimization Based on Video Recognition

作　　者：常铮 李晓艺[1] 刘丹丹 王宪彬[1] CHANG Zheng;LI Xiao-yi;LIU Dan-dan;WANG Xian-bin(School of Transport and Transportation,Northeast Forestry University,Harbin 150040 China)

机构地区：[1]东北林业大学交通学院,黑龙江哈尔滨150040

出　　处：《自动化技术与应用》2018年第11期139-142,共4页Techniques of Automation and Applications

摘　　要：随着汽车保有量的增加,道路交通流量持续增长,交叉口的交通压力日益增大,目前国内绝大多数交通信号灯为固定周期控制,这种定时控制算法显然不能根据交通流的变化而及时做出调整,很可能出现绿灯方向无车辆通过,而红灯方向却有大量车辆等待的现象,造成大量交通资源的浪费。因此建立智能交通灯信号配时系统是十分有必要的,本研究采用了视频识别的方法对交叉口车辆数进行实时获取,并将获取到的数据通过优化算法进行处理,对交通信号进行合理的动态配时,起到了减少交叉口交通延误、降低车辆排队时间、减少交通拥堵的作用。With the increase of car ownership, the traffc fow of the road continues to grow, and the traffc pressure at intersections increases. At present, most Chinese traffc lights are controlled at a fxed period. This timing control algorithm cannot change with traffc fow. Thus there may be a phenomenon that no vehicle in the direction of green light, and a large number of vehicles waiting in the direction of the red light, this greatly reduced the traffc capacity of roads. Therefore, it is necessary to establish a signal timing system for intelligent traffc lights. In this study, a video recognition method is adopted to obtain real-time traffic flow at intersections, and the acquired data is processed through an optimization algorithm to make traffc signals become reasonable. The dynamic timing achieves the effect of reducing traffc delays at intersections, reducing vehicle queuing time and traffc congestion.

关 键 词：图像处理 交通信号灯 智能控制系统 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] U491.5[自动化与计算机技术—计算机科学与技术]