基于OpenCV的校园车牌识别计费管理系统毕业论文
2021-12-21 21:46:37
论文总字数:25023字
摘 要
近年来,图象处理技术的发展,为长期以来人力模式下的车辆管理带来了非常明显的好处,车辆识别的自动化已成为交通管理的趋势。车牌识别是我们当下社会对交通进行管理的重要部分,它的核心是图像处理,对包含车牌的车辆图像进行处理,得到我们想要的汽车的车牌号码。当前,车牌识别已被应用于停车场、道路管理、车辆出入等各种地方,对于我们维护交通秩序,防止道路阻塞,实现道路高效管理有着现实的意义。
本课题车牌识别算法核心的模块主要包括三部分,一是用于找到车牌的方法,二是用于字符分割的方法,三部是用于字符识别的方法。本课题还会在车牌识别算法的基础上,通过Qt工具制作一个简易的计费管理系统,能够对校园出入的车辆进行智能识别与管理,该系统包含的主要功能有:用户登录注册、从本地或摄像头获取图片、车位管理、判断是否为校内车并对临时车收费。
课题选取了120张较为清晰的网络图片和在户外拍摄的30张图片作为测试样本,车牌定位的成功率为90%,成功定位的车牌的识别率为85%,总体识别率为76%,识别效果良好。管理系统可以实现正常的登录与注册操作,可以从本地或摄像头获取图像,调用车牌识别算法进行识别,并通过对数据库操作完成管理功能。
关键词:车牌 识别 定位 分割 OpenCV Qt
License plate recognition billing management system based on OpenCV
Abstract
In recent years, the development of image processing technology has brought obvious benefits to the vehicle management under the manpower mode for a long time, and the automation of vehicle identification has become the trend of traffic management.License plate recognition is an important part of traffic management in our current society. Its core is image processing, which can process the vehicle image containing the license plate to obtain the license plate number of the car we want.At present, license plate recognition has been used in parking lots, road management, vehicle access and other places, for us to maintain traffic order, prevent road congestion, to achieve efficient road management has a practical significance.
The core module of the license plate recognition algorithm in this topic mainly includes three parts: one is the method for finding the license plate, the other is the method for character segmentation, and the third part is the method for character recognition.This topic will be on the basis of the license plate recognition algorithm, through Qt tools to make a simple accounting management system, to be able to access the campus of intelligent recognition and management, the system includes the main features are: user login registration, image was obtained from the local or camera, car parking management, determine whether to campus and the temporary car charge.
The project selected 120 clear network pictures and 30 pictures taken in the outdoors as test samples. The success rate of license plate positioning was 90%, the recognition rate of successfully positioned license plates was 85%, and the overall recognition rate was 76%. The recognition effect was good.The management system can achieve normal login and registration operations, can obtain images from the local or camera, call the license plate recognition algorithm for recognition, and complete the management function through the database operation.
Key words: license plate recognition positioning segmentation OpenCv Qt
目 录
摘要 I
Abstract II
第一章 绪论 1
1.1 课题研究背景 1
1.2 课题研究的内容及意义 1
1.3 本文的章节设计 2
第二章 图像处理技术和开发工具 4
2.1 基于OpenCV图像处理技术 4
2.2 Visual Studio 2015 (C ) 4
2.3 Qt 5
2.4 环境配置 5
第三章 车辆牌照定位 9
3.1 RGB与HSV色彩空间对比 9
3.2 基于HSV颜色识别的车牌色彩特征提取 10
3.3 车牌筛选判断及截取 12
3.4 车牌的倾斜检测和矫正 16
3.5 无效边缘去除 17
第四章 车牌字符分割 18
4.1 车牌图像预处理 18
4.1.1 车牌尺寸归一化 18
4.1.2 二次边缘去除 18
4.2 车牌字符分割 19
第五章 车牌字符识别 21
5.1 模板匹配算法介绍 21
5.2 基于模板匹配的字符识别算法 23
5.3 二次识别 26
第六章 基于Qt的车位计费管理系统 29
6.1 数据库设计 29
6.2 登录与注册界面 31
6.3 选择获取图像方式界面 34
6.4 车位计费管理界面 37
第七章 结论和展望 41
参考文献 43
致谢 42
绪论
课题研究背景
近年来,图象处理技术的发展,为长期以来人力模式下的车辆管理带来了非常明显的好处,车辆识别的自动化已成为交通管理的趋势。车牌识别是我们当下社会对交通进行管理的重要部分,它的核心是图像处理,对包含车牌的车辆图像进行处理,得到我们想要的汽车的车牌号码。
请支付后下载全文,论文总字数:25023字