计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (20): 188-190.

• 图形、图像、模式识别 • 上一篇    下一篇

一种基于新特征的车牌检测方法

江林升1,朱学芳2   

  1. 1.南京森林警察学院 信息系,南京 210046
    2.南京大学 信息管理系,南京 210093
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-07-11 发布日期:2011-07-11

Car plate detection based on new features

JIANG Linsheng1,ZHU Xuefang2   

  1. 1.Department of Information,Nanjing Forest Police College,Nanjing 210046,China
    2.Department of Information Management,Nanjing University,Nanjing 210093,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-07-11 Published:2011-07-11

摘要: 车牌检测是车牌识别的关键所在,两种新的区域统计学特征能迅速排除大量的非车牌区域,在此基础上,采用增加了辅助判决的级联分类器来改进AdaBoost算法。实验表明,该算法与基于颜色特征分类器和传统的级联AdaBoost分类器相比,具有较快的检测速度、较高的检测率和较低的误检率。

关键词: 特征, 级联分类器, AdaBoost

Abstract: License plate detection is the key to the recognition of automatic license plates.The two new regional statistics can quickly rule out a large number of non-regional plate vehicles,based on which,a noval Adaboost with auxiliary detection is proposed in the paper.Tests show that compared with color-based classifier and traditional cascaded AdaBoost classifier,the new algorithm takes the advantages of a faster detection speed,a higher detection rate as well as fewer false detections.

Key words: feature, cascaded classifier, AdaBoost