摘要车牌识别技术是指对摄像机采集到的照片进行处理[1]，自动提取车牌信息（包括阿拉伯数字、英文字母、汉字）的技术，从而获得每辆车所特有的车牌，车牌识别技术极大地提高了交通管理效率，越来越引起了人们的重视。80575

车牌识别的第一步是车牌定位，本文首先对图像进行灰度化处理来提高处理速度并且减少存储空间[2]，接着用roberts算子进行边缘检测使得车牌区域更为清晰，然后进行形态学处理从而显示出可能为车牌的连通区域，最后通过对图像中像素点分析定位到车牌。

车牌识别的第二步是车牌字符分割，本文先对定位到的车牌图像进行二值化处理，再进行形态学滤波去除噪声和不必要的区域，再通过对图像进行垂直投影分析，从而确定了每个字符的位置，再对每个字符进行切割。

车牌识别的第三步是分割字符识别，本文通过模版匹配法来识别分割字符，通过将分割出来的字符与数字、字母和汉字模版相比较，选出差距最小的模版即为结果。

本文应用matlab编程，实现了对车辆图像的车牌定位、分割和字符识别，完成了车牌识别工作。

毕业论文关键词  车牌识别 车牌定位 字符分割 字符识别 matlab

毕业设计说明书外文摘要

Title  The research of license plate recognition technology

Abstract License plate recognition technology is the technology which can extract license plate information automatically,it can deal with the picture collected by video camera so that we can get the unique license plate information from every car,which have improved the efficiency of Traffic Management greatly。And it has aroused more an more people's attention。

The first step of license plate recognition is the location of license plate。first,we make the image to a gray one to reduce the scale of data processing and improve the data processing speed。then,we use roberts operator to make the license plate area more clear。Afterwards, we carry on morphological processing to show the possible area of license plate。At last,we locate the license position through the analysis of pixel in the image。

The second step of license plate recognition is character segmentation of license plate。First,we do binarization processing to license plate。Then we remove the noise and the unnecessary area。Next,we do vertical projection analysis to the image so that wo can confirm the location of every character。Finally,wo can easily segment each character。

The third step of license plate recognition is character recognition。We use template matching to recognize the character。We have to compare the segmented character to the template characters to get the most accurate result。

Keywords  matlab;License plate recognition;location of license plate;

character segmentation;character recognition

目  次

1  绪论 1

1。1  车牌识别技术的研究背景与研究意义 1

1。2  车牌识别研究的国内外现状 1

1。3  车牌识别流程 1

1。4  本文主要研究内容及章节安排 2

2  车牌定位 4

2。1  灰度化 4

2。2  边缘检测 5