毕业论文--车牌自动识别技术的研究与实现.doc

湖南大学毕业设计(论文) 第 PAGE 3页 HUNAN 毕 业 论 文 论文题目 车牌自动识别技术的 研究与实现 学生姓名 学生学号 专业班级 电子1102 学院名称 电气与信息工程学院 指导老师 学院院长 2015 年5 月 湖南大学毕业设计(论文) 第 PAGE I页 摘 要 车牌自动识别系统是智能交通系统的重要组成部分之一。我国汽车总量较大，对交通管理系统的性能要求较高，车牌自动识别系统在具有巨大的实用价值和广阔的发展前景 本文首先对目前车牌识别系统的算法及原理进行了分析，之后结合我国常见牌照的格式规范，提出了一种以MATLAB软件为基础的车牌自动识别算法。该算法主要由三大主要部分组成，分别实现牌照定位、字符分割、字符识别与获取功能。在对车辆图像进行预处理之后，牌照定位部分使用边缘提取、区域标记等方法对牌照区域进行定位；字符分割部分利用投影法实现分割功能；字符识别部分使用模板匹配法对牌照字符进行识别、获取。仿真结果表明，该算法对国内常见牌照具有较为理想的识别效果，达到了预期设计标准。 关键词：MATLAB；车牌识别；投影法；模板匹配；边缘提取；区域标记 Research and Implementation of Vehicle License Plate Recognition Abstract License Plate Recognition (LPR) System has become one of the important parts of the Intelligence Transportation System (ITS), the study of which has positive effects on developing the ITS, increasing the living standard of the general public, etc.. A lot of researches have already been carried out domestic and foreign, and the researchers have proposed lots of practical (or creative) methods. In the future, new technologies and algorithms will be applied on the LPR System, of which the degree of integration, the scope of application, reliability, etc. will rise greatly. There has been a large amount of automobile and higher demand of the performance of traffic management system in China, thus LPR System has great value and broad prospects for development. Firstly, the paper analyzed the algorithm and principle of the existing LPR System; then a kind of LPR algorithm is put forward based on MATLAB, which consists of three main parts: License Plate Location, Character Segmentation, Character Recognition and Acquisition. After the pretreatment of the vehicle image, the License Plate Location part locates the license plate region by means of edge detection, regional labeling, etc.; Character Segmentation part uses projection method to complete separation; Character Recognition and Acquisition part recognizes the character and obtains them using template matching method.