基于ARM的车牌识别系统界面毕业设计论文.doc

大连海事大学 毕 业 论 文 二○一四年六月 基于ARM的车牌识别系统界面设计 专业班级：通信工程2班 姓 名： 雷 良 指导教师： 谭 克 俊 信息科学技术学院摘 要 车牌识别技术(VLPR)作为智能交通系统（ITS）的重要方面，其和蓬勃发展的嵌入式系统相结合，有着重要的理论意义和广阔的应用前景。本文在实现了对嵌入式Linux图形用户界面设计的基础上，重点研究了车牌识别的各项关键技术，包括图像预处理、车牌定位、字符切割和字符识别。 本文首先对车牌识别系统的开发环境和嵌入式图形用户界面的设计方面进行了介绍，开发环境是基于OpenCV的Qt开发环境，在Qt下设计并实现了用户自定义的图形化人机接口界面。 本文利用C++语言，通过调用OpenCV库函数实现了车牌识别中图像预处理、车牌定位、字符切割和字符识别等各种算法。在图像预处理方面，本文在对大量图片进行了综合处理的基础上，比较了各种预处理方法，最后采用基于HSV阈值化后进行滤波及形态学处理的预处理方法，该方法能很好的消除图形噪音。在定位方面，本文提出轮廓并分级的策略 关键词：Qt；图像预处理；车牌定位；字符分割；字符识别 ABSTRACT License Plate number Recognition Technology (VLPR) as an important aspect of intelligent transportation systems (ITS), It is combining with the rapid growth of embedded systems, In addition that has the important significance and broad application prospects. This article is focuses on some key technologies of license plate number recognition: include the image preprocessing and the license plate positioning and the character segmentation and character recognition, They are based on the design of Embedded Linux Graphic User Interface.  Firstly introduce the design of Embedded Linux Graphic User Interface and the development environment of License Plate number Recognition., The development environment is based on the Qt of OpenCV, Then designing and implementing a graphical human machine interfaces of user-defined on the Qt. In the thesis, by Used C + + language, then adjust using the OpenCV library functions to got some algorithms of the image preprocessing and the license plate positioning and the character segmentation and character recognition. For the image preprocessing through more images manipulation and compare of various pretreatment methods. Then through Pretreatment method of the filter and the morphological processing based on a HSV threshed image, This method is better to eliminate the graphical noise. For license plate positioning that I’m make a strategy of contour detection and Classification judgment. For character segmentation th