基于boosting算法的交通事件检测  

Traffic Incidents Detection Based on Boosting Method

作　　者：孙熙[1] 李夏苗[1] 

机构地区：[1]中南大学交通运输工程学院,长沙410075

出　　处：《交通运输系统工程与信息》2007年第5期37-41,共5页Journal of Transportation Systems Engineering and Information Technology

基　　金：国家自然科学基金(50378091)

摘　　要：提出一种新颖的基于boosting模糊分类的交通事件检测方法.该方法利用Boosting方法和遗传算法以迭代形式获取一组模糊规则及规则对应的权值,分类器以加权投票方式进行分类决策.运用Matlab进行了仿真分析,结果表明提出的交通事件检测算法利用较少样本数据即可快速实现交通事件检测.A novel method is proposed for traffic incidents detection based on boosting fuzzy classification. The fuzzy rule and the weight have been obtained using boosting method and Genetic algorithm in an incremental fashion. The classified using aggregate the castled votes to carry through the classified decision-making. Finally we have simulated with Mat lab and found out that the algorithm can detect incidents fleetly using a few samples.

关 键 词：交通事件检测 遗传算法 BOOSTING方法 模糊分类器 

分 类 号：U491[交通运输工程—交通运输规划与管理]