摘要： 数字识别的研究和在各个方面的应用有很多，可以说是十分的广泛。而且有的时候对识别率的要求很高，特别是一些要求非常严谨的领域。本文是对将已经切分出来的数字用MATLAB进行识别，由于识别的是邮编，邮编的格式现在发展的比较多样化，现在的邮编基本是直接打印好的，常见于一些快递包裹和信件。但也存在一些手写的邮编。本次毕业设计用到的识别样本即有手写的数字，又有打印版的数字。在识别的过程中遇到了一些问题，样本中邮编是有红色的边框，去边框就成了一项难题，后来也想办法解决了。接下来就是样本的和数字数据库的比对问题，刚开始在比对的时候，相似度很低，识别效果不是很好，后来经过多次的比对，已经达到很好的识别效果。基本的识别过程是：读取样本；预处理，这里面包含转化灰度图像，二值化处理等；然后就是比对识别。29958
毕业论文关键词： MATLAB；识别；灰度化；二值化；比对
Zip code automatic identification system of development - digit recognition research
Abstract: Digital recognition research and application in all aspects has a lot of, can say is very broad. And sometimes claim to the recognition rate is very high, especially some very rigorous field requirements. This article is to have split out number for identification by MATLAB, as a result of recognition is zip code, zip format now for the development of more persified, the zip code base is printed directly, common in some express packages and letters. But there are also some handwritten zip code. This graduation design is used to identify sample have handwritten Numbers, and there was a print version of the figures. Met with some problems in the process of identifying, samples in the zip code is a red border, border has become a difficult problem. Then think of some way to solve. Next is the problem of sample and digital database comparison, began at the time of comparison, similarity is low, the recognition effect is not very good, later after multiple comparison, has reached a good recognition effect. Basic identification process is: read the sample; Pretreatment, which contains into gray image, binarization; processing, etc.; Then compare identification.
Keywords: MATLAB; recognition; graying; binarization ; comparision
目录
摘要 i
Abstract i
目录 v
1. 绪论 1
1.1 邮编识别产生的背景和意义 1
1.2 国内外研究现状 2
1.3 论文的组织结构 3
2. 识别技术简介和应用 5
2.1 数字图像处理简介 5
2.2 模式识别简介 5
2.3 模式识别方法 6
2.3.1 决策理论方法 6
2.3.2 句法方法 6
2.4 数字图像处理的应用 6
3. MATLAB介绍 8
3.1 MATLAB的概述 8
3.2 MATLAB产生的历史背景 8
3.3 MATLAB的语言特点 9
3.4 MATLAB在图像处理中的应用 10
4. 课题的分析和设计 12
4.1 图像相关理论分析 12
4.1.1 图像分析 12
4.1.2 图像的读取 12
4.1.3 图像预处理 13 MATLAB邮编自动识别系统的开发数字识别研究:http://www.751com.cn/zidonghua/lunwen_25416.html