汽车牌照自动识别系统的设计和研究  

Research on License Plate Recognition System Design

作　　者：张涛[1] 齐永奇[1] 

机构地区：[1]华北水利水电学院机械学院,河南郑州450011

出　　处：《华北水利水电学院学报》2011年第4期19-21,共3页North China Institute of Water Conservancy and Hydroelectric Power

摘　　要：车牌识别是指通过图像处理、模式识别和统计分析等方法从车辆图像中提取车牌字符信息,从而确定车辆身份的技术.车牌识别包括图像采集、图像预处理、车牌定位、车牌分割、字符识别等5个核心部分.这里提出一种基于MATLAB的车牌识别系统的实现方法,综合使用多种方法提高系统的有效识别能力.该方法可以解决在有噪声和光照恶劣情况下车牌定位、字符倾斜和字符提取等复杂问题.实验结果表明,所述方法是有效的,能够定位和识别车牌字符,达到了预期目的.License plate recognition （LPR） is the technique of draw the character information from the vehicle image to confirm the i- dentity of the vehicle through using image manipulation, pattern recognition, statistical analysis, and so on. LPR consists of five steps ： image gathering, image pre-proeessing, license plate location, character segmentation and character recognition. A method is proposed to realize the LPR based on MATLAB, and comprehensively uses a number of ways to improve effective recognition of the system. The method solves the complex problem of license plate location, license plate tilt, and character extraction in the bad ease of the noise and light conditions. The experimental results show that this system ean work effectively, locate and identify license plate number. It can achieve the desired design purpose.

关 键 词：车牌识别 图像处理 车牌定位 字符识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]