基于特征匹配算法的车型识别研究  

The Study of Vehicle type Recognition Arithmetic Based on Feature Matching

作　　者：杨文强[1] 吴冰[1] 陈步威[1] 

机构地区：[1]河南理工大学电气工程与自动化学院,焦作454003

出　　处：《信息通信》2008年第4期73-75,共3页Information & Communications

基　　金：河南省自然科学基金资助项目(0211060500);河南省攻关项目资助(0124140156)

摘　　要：随着机动车数量的迅速增加,交通管理的难度日益加大与日俱增。毋庸置疑,汽车车型自动识别技术在实现交通领域智能化方面将占据举足轻重的地位,有着深远的现实意义,本文提出了一种基于特征匹配的汽车车型识别算法,实验证明该算法不仅计算量小、稳定性及准确性高、而且鲁棒性好,所以能从根本上满足识别实时性的要求,从而汽车车型的识别率在很大程度上得到了提高,其应用前景将十分广阔。motor vehicles are mushroomming,along with which the difficulty of traffic management is increasing. undoubtedly, the relization of modern traffic management＇s intelligentize will largely depend on vehicle type automatic recognization, which has far-reaching and practical meaning, vehicle type Recognition Arithmetic Based on Feature Matching is proposed in this paper,the experiment indicates that the arithmetic not only decreases calculation, high stabilization and veracity, but also good robust, thus it can radically satisfy the request of real-time, furthermore, identify percentage of vehicle type recognition is greatly enhanced, its application foreground is very expansive.

关 键 词：车型识别 特征匹配 特征提取 二值化 边缘检测 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]