用U描述子识别数字 毕业设计.doc

毕 业 论 文 （科 学 研 究 报 告） 题 目 用U描述子识别数字2009 学 号 0911117027 姓 名 指导老师 教授 华 侨 大 学 教 务 处 2013年6月 摘 要 数字识别系统作为字符识别系统的基础，能够准确地、快速地将数字识别出来。同时也为研究其它字符提供了探索的原理和方法。 本文提出了一个相对较完整的数字字符识别方法，主要将对提取边缘点和边缘点单向单点搜索进行主要论述。其它内容还包括综述了国内外流行的字符识别算法，并相应介绍如何基于VC++实现系统的功能。同时也将对灰度均值二值化算法、边缘提取、算法应用、模板建立等相关过程进行论述。由于本文采用的方法是单点单方向法对数字进行边缘搜索，故会出现几个边缘搜索的几个特殊情况，这些特殊情况的剔除将大大提高系统的准确性，再利用模板匹配法进行数字字符的识别。 数字字符识别系统的实现效果表明，本文介绍的数字字符识别方法速度快，适合应用在实时系统中，适应性较强，识别率也不错。 关键词：数字识别； 灰度化； 二值化； 边缘点 ABSTRACT Number recognition system as the basis of the character recognition system, it is possible to be identified the number accurately and quickly.And it given an Exploring the principles and methods to study other characters. This paper presents a relatively complete?numeric character recognition methods, and will discusses the major which including extracting the edge points and the method of single-direction-single-point’edge-points search. Other also includes the popular at home and abroad character recognition algorithm and describes how to achieve the function of the system and the corresponding based on VC + + . Also the gray binarization algorithm, edge extraction algorithm, template establishment processes will be discussed. In this paper, the method is a single point in one direction on the numeric edge search, so there will be several edge search a few special circumstances, these special circumstances culling will greatly enhance the accuracy of the system, and then use the template matching method for digital character recognition. Achieve the effect of the digital character recognition system indicate that this article describes the digital character recognition method is fast and suitable for application in real-time systems, adaptability, the recognition rate is also good. KEY WORDS:Digital identification; Gray; Binarization; Edge points 目 录 前 言 ….... 1 第1章 基于VC++的数字识别系统的基本框架 2 1.1 VC++简介 2 1.2 基于VC++系统框架的建立 2 1.2.1 建立单文档窗口 2 1.2.2 添加C