一种基于纹理特征和轮廓分析的车牌定位方法  

A License Plate Location Method Based on Texture Feature and Contour Analysis

作　　者：吕文强 朱贤平 Lv Wen-qiang;Zhu Xian-ping(The 28^th Research Institute of China Electronics Technology Group Corporation,Jiangsu Nanjing 210007)

机构地区：[1]中国电子科技集团公司第二十八研究所,江苏南京210007

出　　处：《电子质量》2020年第12期3-6,共4页Electronics Quality

摘　　要：针对车牌识别系统中由于车牌定位率低而影响识别率的问题,该文提出了一种新型的基于图像纹理特征和轮廓分析的车牌定位方法。该方法首先对原始图像进行灰度化和对比度增强处理,然后运用相邻灰度级像素数目统计的方法对图像进行二值化,最后通过纹理特征以及轮廓分析的方法实现车牌的精确定位。In order to solve the problem of low license plate localization rate in license plate recognition system,a new license plate localization method based on image texture feature and contour analysis is proposed in this paper.In this method,the original image is firstly grayed and contrast enhanced,then the image is binarized by counting the number of pixels in adjacent gray levels,and finally the precise location of license plate is realized by the method of texture feature and contour analysis.

关 键 词：车牌定位 图像二值化 纹理特征 轮廓分析 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]