基于支持向量机方法的车型分类技术  

Vehicle Classification Technique Based on Support Vector Machine

作　　者：葛威 王飞[2] 

机构地区：[1]海军驻沈阳地区电子系统军事代表室,辽宁省沈阳市110003 [2]海军驻南京地区电子设备军事代表室,江苏省南京市210013

出　　处：《电子工程师》2008年第4期70-74,共5页Electronic Engineer

摘　　要：车型分类是交通流检测系统的子功能,也是ITS(智能交通系统)中的重要环节。支持向量机方法被看做是对传统学习分类方法的一个较好的替代,特别在小样本、非线性情况下,具有较好的泛化性能。文中基于视频检测技术,采用支持向量机方法对车型分类进行了研究。实验表明,支持向量机方法是一种很有前景的车型分类技术。Vehicle important part of the In Classification technique is a sub-function of the vehicle detection system, is also an telligence Transportation System （ITS）. Support Vector Machine approach is considered a good candidate because of its good generalization performance, especially when the number of training samples is very small and input space is nonlinear. This paper presents a study on the vehicle classification based on support vector machine. Experimental results show that the classification performance of support vector machine is a promising technique for vehicle classification.

关 键 词：车型分类 交通流检测系统 支持向量机 

分 类 号：TP274.3[自动化与计算机技术—检测技术与自动化装置]