一种基于相似度的汽车车牌号码自动识别系统  

A Likelihood-based Recognition Method for A Vehicle Number Plate

作　　者：郑宇[1] 肖南峰[1] 

机构地区：[1]华南理工大学,广州510641

出　　处：《交通与计算机》2004年第5期21-25,共5页Computer and Communications

基　　金：国家自然科学基金资助 ( 60 3 75 0 3 1 );广东省自然科学基金资助 ( 3 65 5 2 )

摘　　要：文章介绍了一种基于相似度的汽车车牌号码自动识别系统。介绍了系统的软、硬件设计 ,分析了车牌图像的预处理、车牌定位、车牌区域二值化和车牌字符识别。实验结果表明 。In order to automatically recognize a vehicle number plate, this paper introduces an auto-recognition system for a vehicle number plate based on likelihood. The system can recognize not only the blurry vehicle plate image, but also the incomplete and lean image. It can also specify the vehicle identity on time by contrasting the recognition result in the vehicle number plate database. The likelihood-based system shows high stability by the experimental results. The system can be widely used in monitoring entrances of a building, airports, stations and so on.

关 键 词：相似度 车牌图像 车牌字符识别 自动识别系统 车牌定位 二值化 车牌号码 汽车车牌 硬件设计 稳定性 

分 类 号：U492.22[交通运输工程—交通运输规划与管理] TP391[交通运输工程—道路与铁道工程]