基于颜色与结构特征的车牌定位算法  

Algorithm for license plate locating base on the feature of color and structure

作　　者：高一文[1,2] 龚劬[1] 

机构地区：[1]重庆大学数学与统计学院,重庆沙坪坝400030 [2]重庆文理学院教务处,重庆永川402160

出　　处：《重庆文理学院学报（自然科学版）》2012年第1期71-75,共5页Journal of Chongqing University of Arts and Sciences

摘　　要：针对车牌具有稳定的颜色特征和形状特征,提出基于颜色与结构特征的车牌定位算法.利用OTSU自动阈值化技术将灰度车牌图像转化为二值化图像,通过对每一个连通区域提取形状参数,粗划分出候选车牌区域,最后利用颜色特征确定车牌.实验表明,这种车牌定位方法具有一定的优越性.In terms of the stable feature of color and structure of vehicle license plate,the algorithm for license plate locating based on the feature of color and structure was proposed.The OTSU automatic threshold technology was used to turn the gray image into binary image.By extracting shape parameter of each connected region,the area of license plate was roughly drawn out,and the license plate was located according to the color feature.The experiment proves that the method has certain advantages.

关 键 词：车牌定位 颜色特征 结构特征 二值化图像 

分 类 号：TP391.413[自动化与计算机技术—计算机应用技术]