基于51单片机的高速公路测速系统和车牌识别分析
添加时间: 2010-3-20 11:19:19 文章来源: 文章作者: 点击数:17688
摘 要
鉴于高速公路限速牌不能很好地对司机起到警示作用的作用,本文设计了一套基于MCS-51单片机,包含光电探测装置和显示装置的电子屏幕。它不仅能方便设置并显示该路段的限制速度,以完成普通电子限速牌的限速提示功能,同时能将测得的车速实时显示,并自动判断是否超速。另外它低廉的造价和经计算证明较高的精度大大提高了它的可用性。
车辆牌照自动识别系统是近几年发展起来的基于图像和字符识别术的智能化交通管理系统,是目前国内外模式识别应用研究领域的一个热点。本文对系统中区域提取、图像预处理、字符分割和字符识别等环节涉及的算法、设计做了一个比较详细的论述。本文在图像预处理中重点介绍一种在图像获取阶段有目的定位关注的物体,讨论了灰度图像二值化的多种算法,利用它在原始图像形成的标识区域特性,在约束条件下,按照识别牌几何特征提出了一种特殊的二值化处理方法。实验证明该图像识别系统具有较高的可靠性与稳定性,减小了进一步车牌识别中计算量大的问题,从而提高了车牌识别的准确性和快速性。讨论了灰度图像二值化的多种算法基于数学形态学的图像去除噪声的方法。基于数学形态学的图像去除噪声是通过对图像的开、闭操作有选择的去噪声。可以去除直径小于字符笔划半径的孤立噪声点。还详细地介绍了基于字符形态划分的字符识别方法。基于字符形态划分的字符识别方法是在对数字字符结构进行充分分析的基础上,对基元检测,归纳字符形态特征,得到的快速字符识别方法。
关键词 :光电检测;车速测量;单片机;电子限速牌;车辆牌照;图像处理;基元检测;字符识别
基于51单片机的高速公路测速系统和车牌识别分析
Abstract
In view of the highway speed limit unlicensed drivers should not very well serve as a warning to the role, this article has designed a single-chip based on the MCS-51, including the photoelectric detection devices and display devices of the electronic screen. It not only can easily set up and display the road speed limit in order to achieve common electronic speed limit signs prompt function, can be simultaneously measured real-time display of speed and automatically determine whether the speeding. In addition it is low cost and the higher the accuracy of calculations greatly enhance its usability. Vehicle License Plate Recognition system is developed in recent years based on the image and character recognition operation of the Intelligent Traffic Management System, the application of pattern recognition at home and abroad are currently a hot area of
research. In this paper, the system of regional extraction, image preprocessing, character segmentation and character recognition algorithm, such as aspect involved in the design to do a more detailed exposition. In this paper image pre-processing in the introduction of a focus at image acquisition phase has the purpose of positioning objects of concern, discussed the gray image binarization of a variety of algorithms, use it in the original logo image formation of regional characteristics, in binding conditions, identification card in accordance with the geometric characteristics of a particular binarization approach. Experiments prove that the image recognition system has high reliability and stability, further reduce the vehicle license plate recognition in the calculation of a large quantity of questions, thereby increasing the accuracy of license plate recognition and speed.Discussed the gray image binarization algorithm of multiple images based on mathematical morphology method to remove noise. Images based on mathematical morphology to remove the image noise is through the open and close operation has chosen to noise. Can remove the character strokes of a diameter less than the radius of the isolated noise points. Also detail the division of character-based form of character recognition methods. Morphological character-based division of Character Recognition on the figure are at a full analysis of character structure based on element detection, morphological characteristics summarized characters get Character Recognition Express.
Key words: Photoelectric detection; speed measurement; Singlechip; electron speed licensing; vehicle license; image processing; motif detection; Character Recognition
目 录
摘
要............................................................................................................ I
Abstract....................................................................................................... II 第1章 绪
论........................................................................................... 2
1.1 车牌字符识别研究课题的背
景...................................................... 2 1.2 车牌字符识别研究的意
义.............................................................. 2
1.3 车牌字符识别研究的应用现状及发展.......................................... 3 1.4 本文主要内
容.................................................................................
. 4
第2章 车牌图像预处
理........................................................................ 5 2.1 数字图像处理的相关介
绍............................................................. 5 2.1.1 数字图像处理概
念....................................................................... 5 2.1.2 图像的数字化表
示....................................................................... 5 2.1.3 本文中图像处理所涉及的相关领域........................................... 6 2.2 图像二值
化..................................................................................... 6
2.2.1 彩色图像和灰度图
像................................................................... 6 2.2.2 基于灰度的图像二值
化............................................................... 7 2.2.3 图像二值化结果演
示................................................................... 9
2.3 用数学形态学的方法去除噪声.................................................... 10
2.3.1 数学形态学的几种基本运算..................................................... 10
2.3.2 经开闭运算前后的图像对比显示............................................. 16 2.4 单个字符图像的分
割................................................................... 17 2.4.1 我国车牌的特
点......................................................................... 17 2.4.2 对所要识别的车牌的分
析......................................................... 17
2.4.3 基于列扫描黑色像素积累的字符分割..................................... 18 2.4.4 已经分割后的单个字符演示..................................................... 18 2.5 本章小
结........................................................................................ 19
第3章 基于字符形态划分的字符识别............................................. 19 3.1 字符识别概
述................................................................................
19
3.1.1 目前字符识别的一些常规方法................................................. 20
3.2 字符轮廓分
析................................................................................ 21
3.2.1 字符轮廓的划
分......................................................................... 21 3.2.2 字符四方向轮廓公式化表示..................................................... 21
3.3 字符轮廓的变化特
征.................................................................... 22 3.4 字符形态划分的结构基
元............................................................ 22 3.4.1 字符结构基元划分原
理............................................................. 22 3.4.2 字符形态划分方法的特
点......................................................... 23
3.5 利用字符结构基元划分的字符识别原理.................................... 24 3.5.1 基元的检
测................................................................................. 24
3.5.2 轮廓的统计特
征......................................................................... 25 3.5.3 用字符形态识别数字字
符......................................................... 25 3.5.3.1 数字字符的特
点...................................................................... 25 3.5.3.2 数字字符的识别方
法.............................................................. 26
3.5.4 字符识别的matlab算法流程.................................................. 27 3.6 本课题整体流
程............................................................................ 29 3.6.1 对本课题流程的总体说
明......................................................... 29
3.6.2 用MATLAB完成本课题的流程图........................................... 30 3.7 基于字符形态划分的字符识别方法的特点和不足.................... 30 3.8 本章小
节.................................................................................
....... 31 结
论.......................................................................................................... 32 参考文
献................................................................................................... 33
第1章 绪 论
1.1 车牌字符识别研究课题的背景
随着21世纪经济全球化和信息时代的到来,作为信息来源的自动检测、图像识别技术越来越受到人们的重视。近年来计算机的飞速发展和数字图像技术的日趋成熟,为传统的交通管理带来巨大转变,先进的计算机处理技术,不但可以将人力从繁琐的人工观察,监测中解放出来,而且能够大大提高其精确度,汽车牌照自动识别系统就是在这样的背景与目的下进行开发的。汽车牌照等相关信息的自动采集和管理对于交通车辆管理、园区车辆管理、停车场管理、交警稽查等方面有着十分重要的意义,成为信息处理技术的一项重要研究课题。
关于车牌识别技术及定位系统研究,在我国已经有了十几年的发展历程,目前系统的应用还处于起步阶段,大规模投入使用的成熟系统还没有出现,汽车牌照识别系统作为改进交通管理的有效工具,技术水平仍需完善。
国内外学者对此已经有了较多工作,但实际效果并不理想,尤其是对车牌自适应性强、速度快、准确率高的高速车牌定位方法还有待进一步研究。另外,对辅助光源要求高,也很难有效解决复杂背景下多车牌移动识别的技术难题,如:车牌图像的倾斜、车牌表面污秽或磨损、光线干扰等都会影响定位的准确性。传统车牌识别一般仅支持单一车辆,背景比较简单。而当今许多实际应用场合,如在繁忙交通路口临时对欠税费、报废、挂失等车辆的稽查,则监视区域比较复杂,现有识别方法无法直接应用;而且多数情况下,同时出现多辆汽车,背景有广告牌、树木、建筑物、斑马线以及各种背景文字等,现有的识别方法也不能很好的适应多变的环境。
本课题主要是在对汽车图像进行细致分析,综合考虑了一定噪声干扰情况的基础上,从实用角度来研究车牌字符分割与识别的算法问题。
[2]
1.2 车牌字符识别研究的意义
智能交通系统(Intelligent Transportation System (ITS))是二十一世纪世界道路交通的发