MATLAB车辆标识识别技术研究
毕业论文关键词：MATLAB;车辆识别；牌照识别；车标；直方图；数据库
Title Study on vehicle logo recognition technology
Abstract： The vehicle identification system is the foundation and an important part in the ITS system.And it has been used in many important part of field.Automatic vehicle identification are demanded in many vital areas, and it has been used very extensive, such as automatic parking charging system, monitoring system for street lamps and high-speed automatic toll collection system and other fields. Automatic recognition of vehicle logo is pided into three steps: image enhancement, image segmentation, mathematical operation and recognition of image. Automatic recognition of vehicle logo can be pided into two aspects: the orientation of the logo and the location of the mark. The license plate location technology both at home and abroad has been very mature, so we have to use based on basic knowledge of license plate can be obtained approximately logo car position. After through a series of image pretreatment and mathematical morphology algorithm, that we can find a approxiamte vehicle logo location. Finally, the histogram of edge detection method and Euclidean distance matching method to complete the sign recognition. The corresponding MATLAB program and processing results are given in the paper.
Keywords：MATLAB; Vehicle recognition ， License plate recognition , Gray histogram and edge histogram ，Database
目录
1绪论1
1.1课题研究背景和意义.1
1.2车标识别的各种方法.1
1.3车标识别的应用.2
2理论分析与算子...4
2.1图像预处理4
2.4车牌定位.8
2.5车辆标志的识别10
2.6本章小结..12
3车标定位13
3.1车标定位的总体设计..13
3.2车标的粗定位14
3.3车辆标志的精确定位...15
3.4本章小结20
4.车标的特征库及模版匹配.21
4.1直方图运算21
4.3本章小结.26
5.实验结果分析与总结.27
5.1灰度图的识别法27
5.2边缘直方图的识别方法.30
5.3实验的不足和改进的设想33
总结35
致谢36
参考文献..37
1 绪论 1.1课题研究背景和意义 车标是车辆的特有特性，人们不会去可以改变和掩盖，为了更好地进行交通管理以及疏导， 我们需要进行车标的识别来获取车辆品牌信息。通过对车标的图像和视频画面进行处理和识别，即可达到要求获得汽车标志信息。同时十几年来，随着现代的人类的生活水平的普遍的提高，科学技术的飞速发展，小轿车已经变得非常普及，我们经常会遇到交通拥堵问题，停车也变成了一个亟待解决的问题。因此如何管理调控出入车辆十分重要。
1.2车标识别的各种方法 车标识别有两个步骤进行完成为包括车标定位和标志识别匹配。完成定位车标之后,车标的识别问题就转化为图像识别问题，即为分割和处理所需识别的图像与并进行车标库的建立并用车标库进行匹配分析。目前主要有以下几种识别方法[4]：
上一篇：图像配准技术研究