【作者】 邓婷；

【导师】 李树涛；

【作者基本信息】 湖南大学 ， 模式识别与智能系统， 2007， 硕士

【摘要】 进入21世纪以来,随着社会的发展和人民生活水平的提高,车辆的数目不断增加,交通状况日趋恶化,给社会和环境到来了巨大的压力。智能交通系统ITS(Intelligence Transport System)是在比较完善的道路设施基础上,将先进的电子技术、信息技术、传感器技术和系统工程技术集成运用于地面交通管理所建立的一种实时、准确、高效、大范围且全方位发挥作用的交通运输管理系统。这种系统能够解决由交通拥挤所产生的各种道路问题,因而越来越受到人们的关注。车牌识别(VLPR,Vehicle License Plate Recognition)技术是智能交通系统中的关键技术之一,同时它的发展也十分迅速,已经逐渐融入到我们的现实生活中。车牌识别系统可以对来往车辆进行自动登记、验证、监管和报警,是现代高速公路收费管理系统、高速公路超速自动化监督系统、公路布控、停车场自动收费管理等领域的重要组成部分。本文对车牌识别系统中的图像预处理、车牌定位技术、字符分割和非汉字字符识别几个方面进行了比较深入的论述。主要工作有以下几个部分:(1)车牌定位技术。文章从介绍基于形态学的车牌定位技术入手,通过试验对比,引入了另外一种基于投影分析的车牌定位方法,实现了车牌的精定位及倾斜校正;(2)字符分割算法。使用车牌的投影特征和车牌字符排列的特点对字符进行了分割;(3)车牌区域非汉字字符识别。利用字符本身的特征进行统计,对字符的各项指标进行运算从而识别出字符。实验结果表明,本文所采用的方法能够达到较好的定位效果,具有一定的鲁棒性和实时性。从中可以看出:多种方法相结合的定位分割技术能够提高系统的整体性能,在有效、实用的前提下将多种算法与人工智能技术优化组合将成为模式识别领域的重要发展方向。更多还原

【Abstract】 In 21 century, with the fantastic spur in economy and rapid development of the owning amount of vehicles, the highway communication becomes more and more deteriorative, which brings lots of pressures to the environment and the society. Intelligent Transportation System is a system which can process the transportation management efficiently on the basis of comparative perfect road infrastructure, combines technetronic、information technology、sensor technology and system engineering technology. It can solve most of the problems, that’s why many scientists pay attention on it.License Plate Recognition (LPR) is one of the critical techniques for ITS. The system can register, verify, monitor the vehicle or report to police with automatic license plate recognition automatically. So it can be used in many kinds of occasions, such as the charges system of expressway, monitoring system at road and roll-gate, changes and monitoring system at the district, parking area, guide system, the system counting the quantity if vehicle passing in a certain period time, etc.This paper accomplished research and design of vehicle LPR algorithm. The system consists of three main modules:(1) Location of vehicle license plate. First it introduces a method which uses morphology to extract the plate, then, it presents another method to extract the plate which uses projection analysis. (2) Character segmentation. It divided the LP image to independent character and the method it adopted uses projection feature and the character’s tincture. (3) Character recognition. It puts emphasis on the non-Chinese characters using character statistics.Based on the experimental results, it proves that this proposed method can relatively locate license plate、segment characters and recognize them, the performance of the system is promising. It indicates that combine feature extractions and the other methods to artificial intelligence under the principle of efficiency and practicality will be major development tendencies of pattern recognition in the future.更多还原