基于小波分析的车牌定位方法  

A License Plate Localization Algorithm Based on Wavelet Analysis

作　　者：沈洋[1] 单士娟[1] 

机构地区：[1]宿迁学院,江苏宿迁223800

出　　处：《中原工学院学报》2010年第5期76-78,共3页Journal of Zhongyuan University of Technology

摘　　要：根据图像中车牌的形态特征和横向纹理属性,运用分形盒维数来描述车牌字符的特征,并结合小波变换来定位车牌,该方法计算简单,且不依赖车牌的颜色、形状、尺寸,具有极好的鲁棒性和实时性.测试结果表明,该方法正确定位率超过93%.According to the morphological characteristics of the image plate and horizontal texture properties,this paper presents a method for vehicle license plate localization using fractal dimension and wavelet analysis.This method is simple and does not rely on the license plate color,shape and size,and has excellent robustness and real-time.Test results show that the rate of more than 93%correct positioning.

关 键 词：小波分析 分形维数 车牌定位 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]