便携式车牌智能识别算法研究与系统毕设论文.doc

淮 阴 工 学 院 毕业设计说明书（论文） 作 者: 学 号： 学 院: 专 业: 题 目: 便携式车牌智能识别算法研究 与系统设计 指导者： (姓 名) (专业技术职务) 评阅者： (姓 名) (专业技术职务) 2015 年 6 月 毕业设计说明书（论文）中文摘要 近年来随着汽车数量的剧增，一个稳定实时的车牌智能识别系统已经成为实现智能化交通管理的关键。本文研究了车牌识别系统的图像预处理、车牌定位、字符分割、字符识别四个模块常用的算法，同时提出了一个用MATLAB编程实现的车牌识别系统方案。本文设计的车牌识别系统采用加权平均的灰度变换和Roberts边缘检测对车牌图像进行了预处理；在车牌定位与分割模块采用边缘检测和形态学运算完成了对车牌的定位；在字符分割模块采用投影分析和字符纹理特征相结合的方法实现了对字符的准确分割；在字符识别模块采用模板匹配和BP网络识别算法完成了车牌字符的识别。实验表明，利用BP网络算法对汉字、字母和数字具有较高的识别率分别可以达到86%、87%、91%。 关键词 图像预处理，车牌定位，字符分割，字符识别 毕业设计说明书（论文）外文摘要 Title The Algorithm Research and System Design of License Plate Intelligent Recognition Abstract In recent years, With the sharp increase in the number of cars，a stable real-time intelligent license plate recognition has become the key to realize intelligent traffic management. This paper mainly discusses and studies about the common algorithms that are used in image preprocessing, license plate localization, character segmentation and character recognition which are four part of license plate recognition system. At the same time, a scheme was proposed which can realize recognizing plate from image by programming in MATLAB. In this system, A weighted average of the gray-scale transformation and Roberts edge detection were used to realize image preprocessing; Edge detection and morphology operations were used to realize Plate localization ; Character segmentation is realized by using projection analysis and character of texture features; Character recognition is realized by using template matching and BP network algorithm . Experiments showed that the system where BP network algorithm is used has a high recognition rate which can reach 86%, 87% and 91% in Chinese characters, English characters and Numbers. Keywords Image Preprocessing, License Plate Localization, Character Segmentation, Character Recognition  目