基于海量图像数据分析的交通协调控制策略框架  

Traffic Coordination Control Strategy Framework Based on Massive Image Data Analysis

作　　者：刘杰 

机构地区：[1]广西职业师范学院计算机与信息工程学院,广西南宁530007

出　　处：《大众科技》2020年第11期1-3,共3页Popular Science & Technology

基　　金：2018年度广西高校中青年教师基础能力提升项目"基于海量图像大数据的城市交通协同调度关键技术研究"(2018KY1272)。

摘　　要：长久以来,交通流识别器在交通管理中扮演重要角色,如何有效利用交通识别视频信息,提升交通区域智能化控制和决策能力,是历来诸多研究者关注的热点问题之一。通过研究并借助现有的图像识别技术,设计了一种可以整合多个路口交通信息的模型框架,该框架具备一定的自学习和自适应能力。实验研究表明,该模型在一定的约束条件下可整合复杂路口信息,并能根据历史信息学习路口状态,预估交通拥堵等关键交通事件,为多路口区域信息智能整合提供了一种较新的研究思路和策略。For a long time,traffic flow recognizers have played an important role in traffic management.How to use traffic recognition video information effectively to improve the intelligent control and decision-making capacity of traffic area is one of the hot issues that many researchers pay attention to.By studying and relying on the existing image recognition technology,a framework that can integrate traffic information of multiple intersections is designed.The framework has certain self-learning and adaptive capabilities.Experimental results show that the model can integrate complex intersection information under certain constraints,and can learn intersection status according to historical information,and predict traffic congestion and other key traffic events.It provides a new research idea and strategy for intelligent integration of multi intersection regional information.

关 键 词：交通控制策略 多路口协同 交通事件检测 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]