基于matlab的车牌字符分割算法研究
2022-12-20 10:27:27
论文总字数:16937字
摘 要
自动车牌识别技术是指通过计算机对已拍摄到的车牌进行识别,从而确定车牌号码的技术。随着现代社会发展的越来越快,汽车保有量持续增加,一些如汽车违法违规事件增多和交通变得越来越拥堵的问题也随之而来。故有专家提出研发出新型的智能交通系统来解决这些问题。而要想研发出行之有效的智能交通系统就必须对其核心组成部分——自动车牌识别系统进行研发。自动车牌识别系统包括三大步骤:车牌定位、字符分割和字符识别。
本文着重研究车牌自动识别中字符分割算法,并且通过matlab2019软件对获取的100副车牌图像原始数据进行编程处理,实现了在车牌自动识别系统中对车牌字符进行分割。字符分割算法的一般步骤是首先对图像进行预处理:包括灰度处理、二值化处理以及对图像进行去噪处理后进行形态学上的开闭运算去除边框等。然后分别使用了两种字符分割的算法:投影法、投影法和固定边界法结合的方法。实验结果表明:在较理想的实验条件下,能够较好的分割出92副车牌,其准确率高达92%,可以满足实验所提出的要求。但是我国的车牌种类、样式繁多,还需深入开展车牌图像识别研究。
关键词:自动车牌识别;灰度化;二值化;字符分割;投影法;投影法和固定边界结合法
Research on Segmentation Algorithm of License Plate Recognition Based on Matlab
Abstract
Automatic license plate recognition technology refers to the technology of recognizing the photographed license plate through the computer, so as to determine the license plate number. As modern society develops faster and faster, car ownership continues to increase, and problems such as increased car violations and increased traffic congestion also follow. Therefore, some experts proposed to develop a new type of intelligent transportation system to solve these problems. In order to develop an effective intelligent transportation system, it must develop its core component, the automatic license plate recognition system. The automatic license plate recognition system includes three major steps: license plate positioning, character segmentation, and character recognition.
This paper focuses on the character segmentation algorithm in automatic license plate recognition.And through matlab2019 software to program the original data of 100 license plate images obtained.Realize the segmentation of the license plate characters in the automatic license plate recognition system.The general steps of the character segmentation algorithm are first to preprocess the image:including grayscale processing,binarization processing,and morphological opening and closing operations to remove the frame after denoising the image.Then two types of character segmentation algorithms are used:projection method,projection method and fixed boundary method.The experimental results show that under ideal experimental conditions,92 pairs of license plates can be segmented well,and the accuracy rate is as high as 92%,which can meet the requirements of the experiment.However,there are many types and styles of license plates in our country,so we need to carry out in-depth research on license plate image recognition.
Key words: Automatic License Plate Recognition,Grayscale,Binarization,Character Segmentation,Projection Method,Projection Method and Fixed Boundary Combination Method
目 录
摘要 I
Abstract II
第一章绪论 1
1.1选题的背景和意义 1
1.2车牌识别概念综述 1
1.3国内外研究现状和发展趋势 1
1.3.1国外研究现状 1
1.3.2国内研究现状 2
1.4我国车牌特征及技术难点 2
1.4.1我国车牌特征 2
1.4.2技术难点 2
1.5本文主要结构 3
第二章车牌图像预处理 4
2.1图像预处理 4
2.2图像灰度化 4
2.3图像去噪 5
2.4灰度变换 6
2.5边缘检测 7
2.6二值化 8
2.7形态学上的处理 9
第三章字符分割 11
3.1引言 11
3.2垂直投影法 11
3.2.1垂直投影法算法流程及实验结果 11
3.3固定边界法 12
3.4投影法和固定边界法相结合的改进方法 12
3.4.1插值算法 12
3.4.2改进法流程及数据结果 14
3.5本章小结 14
第四章总结 15
致谢 16
参考文献 17
第一章 绪 论
1.1选题的背景和意义
随着现代社会的不断发展,经济和科技水平不断的提高,汽车在人们出行生活中也扮演着越来越重要的角色,其保有量也因此不断上升。最近十一年来我国汽车产销量连续蝉联世界第一,甚至有专家看好在2020年末我国汽车保有量有望赶超美国。从经济和消费方面来说这是一件好事,但是从另一方面来看这给我国的交通管理系统也带来了极大的压力。如何更好更有效对公共交通进行管理是我国现阶段面临的一个主要问题。
剩余内容已隐藏,请支付后下载全文,论文总字数:16937字