摘  要
车牌识别技术是智能交通系统中的重要组成部分，它在违章抓拍，不停车收费，停车场管理以及重要场所过往车辆的实时登记等方面都有重要的作用。论文从车牌字符识别的理论出发，基于MATLAB语言对现有的模板匹配，神经网络，基于向量机（SVM）等方法在字符识别过程中的优缺点以及识别率进行系统的研究。论文的主要工作如下：
1.针对车牌图片的预处理包括去噪，增强，分割，提取字符等等；
2.构建模板匹配，神经网络，基于向量机（SVM）字符识别的相关测试数据；
3.分别实现模板匹配字符识别算法，神经网络字符识别算法，基于向量机（SVM）字符识别算法，并做相应识别率的实验，将三者的实验结果进行比对；
4.基于MATLAB GUI做三种算法系统的界面。
关键词：  车牌识别  模板匹配  神经网络  向量机  识别率

ABSTRACT
License plate recognition technology is the intelligent transportation system an important part of it illegal to capture, no parking, parking management, and an important place in the past, real-time vehicle registration and other aspects important role. Papers from the license plate character recognition theory, MATLAB language based on the existing template matching, neural network, based on vector machines (SVM) and other methods in the process of character recognition and the recognition rate of the advantages and disadvantages of the system. The main work is as follows:
1.Pre-treatment, including the license plate image denoising, enhancement, segmentation, extraction of character, etc.
2.Construction of template matching, neural network, based on the vector machine (SVM) t
est data related to character recognition;
>汽车牌照