摘 要

当今世界，互联化已成必然，物联网技术愈来愈成为这次数字革命的焦点。发展成熟且价格低廉的生物识别技术在物联网的普及中发挥着重要的作用，嵌入式终端被要求提供这些技术。出入人员记录、侦探破案等事件中常需要指纹技术的帮助，如今，如手机解锁、电子支付、保险箱验证等指纹识别应用也开始在市场上大放异彩。

生物识别技术是当今世界研究的热点之一。目前，各种诸如视网膜识别、人脸识别、动作识别、声音识别、指纹识别等生物识别技术均飞速发展并被应用于身份认证。其中指纹识别发展最早，由于其低门槛和高稳定性，是目前应用现状和前景较好的生物识别系统。然而，指纹识别在中国乃至世界都不流行。其主要原因是速度慢，性能差，价格高。由于目前国内主要参考国外成熟的指纹识别技术，而相比于本系统中采用的上位机处理，显得笨重、不方便移动。本系统结合了嵌入式技术和指纹识别技术，可以较好解决以上问题。

本文的主要研究工作如下：

1.分析了课题研究的背景、目的及意义，介绍了指纹识别技术的国内外发展概况。研究表明，指纹识别技术具有很强的理论价值，而且具有较高的实用价值和实践意义，发展前景广阔；

2.学习和掌握了指纹识别的主要算法。系统主要是基于指纹的这两种特征来识别：一是组成指纹的整体结构特征；二是指纹的局部的特征点。本系统应用的是一种从指纹的特征点中识别出指纹的算法，通过从细化处理中提取的特征点，形成处理过后指纹图像，将预处理后的指纹图像与模板中的进行匹配，从而识别；

3.学习和掌握了嵌入式指纹识别系统的硬件部分。本系统采用了一款ARM11内核的芯片作为控制核心，指纹采集部分选用光学指纹采集器，本文给出了芯片和指纹采集器的相关介绍及选择原因，并且给出了系统组成框图和接口设计电路；

1. 学习和掌握了嵌入式指纹识别系统的软件部分。本系统主要选用C语言来编程，采用的开发环境为RVDS2.2，并使用ARMCC进行编译，搭建程序运行环境。论文中描述了在S3C6410 芯片上是如何实现指纹识别的，包括图形用户界面的设计。

关键字：ARM；嵌入式；指纹识别；特征匹配

Abstract

In today's world, interconnection is inevitable, the Internet of things technology is more and more becoming the focus of the digital revolution. Mature and cheap biological recognition technology is playing an important role in the popularity of the Internet of things, the embedded terminal are required to provide these technologies, at present, fingerprint identification technology has been applied most widely. Entry personnel records, detectives to solve crimes and other events often need the help of fingerprint technology, now, such as to unlock the phone, electronic payment, insurance box verification application of fingerprint identification also began to shine in the market.

Biometric technology is one of the hot research topics in the world today. At present, a variety of biological recognition technology, such as the rainbow film recognition, face recognition, signature recognition, voice recognition, fingerprint recognition technology has been developed rapidly and used for identity authentication. The development of fingerprint identification is the earliest, because of its low threshold and high stability, it has a better current application status and prospects. Fingerprint recognition, however, is not popular in China and even in the world.The main reason is that the speed is slow, the performance is poor and the price is high. Because the current domestic businesses mainly refer to foreign mature fingerprint identification technology, compared to the system using the host computer processing, it is cumbersome and inconvenient to move.The system can solve the above problems by combining the fingerprint identification technology with embedded technology.

The paper's main research work is as follows：

1.The paper has analyzed the background, purpose and significance of the research, introduced the development of fingerprint identification technology at home and abroad. The research has shown that the fingerprint identification technology had a strong theoretical value, but also had a high practical value and practical significance, there were broad prospects for development；

2.The main algorithm of fingerprint identification has been learned and mastered.The system is mainly based on the fingerprint of these two kinds of features to identify: first, the composition of the overall structure of the fingerprint features; second , the fingerprint of the local characteristics of the point. Application of the system is a kind of from the features of the fingerprint recognition fingerprint algorithm, by feature points extracted from the refinement, forming the processing of fingerprint image and matching the preprocessed fingerprint image with the template to recognize it ;

3.The hardware of embedded fingerprint identification system has been learned and mastered.The system used a kernel ARM11 chip as the control core, and a optical fingerprint collector as the fingerprint sensor, the chip and the fingerprint collector introduction and chosen reasons are given in the paper as well as the system composition block diagram and the design of interface circuit；

4.The software part of embedded fingerprint identification system has been learned and mastered. This system mainly has used the C language to program, usesd the development environment for RVDS2.2, and used the ARMCC to compile, built the program running environment. This paper describes how to implement fingerprint recognition on the S3C6410 chip, including the design of graphical user interface.

Keywords: ARM； Embedded；Fingerprint Identifier；Feature matching

目录

第1章 概述 1

1.1 课题研究的背景及目的、意义 1

1.2 指纹识别技术的国内外研究现状 2

1.3 本文的主要研究内容及组织结构 3

第2章 系统硬件设计方案 5

2.1 指纹识别系统的概述 5

2.2 系统硬件设计 5

2.2.1 总体硬件设计方案 5

2.2.2 指纹采集器介绍及接口设计 6

2.3 本章小结 8

第3章 指纹预处理算法 10

3.1 指纹图片畸变处理算法 10

3.2 指纹图像的分割 11

3.3 指纹均衡 12

3.4 指纹收敛处理 14

3.5 指纹平滑处理 15

3.6 指纹增强处理 16

3.7 指纹的二值化细化处理 17

3.8 本章小结 18

第4章 指纹识别算法 19

4.1 指纹图像的特征 19

4.2 指纹图像特征提取算法 19

4.3 特征点的匹配算法 20

4.4 本章小结 22

第5章 系统的实现与测试 23

5.1 开发环境的设置 23

5.2 指纹算法的验证 26

5.3 指纹采集接口的调试 26

5.4 系统运行的测试 28

5.5 本章小结 29

第6章 总结与展望 30

6.1 总结 30

6.2 展望 30

参考文献 32

致谢 33

第1章 概述

1.1 课题研究的背景及目的、意义

信息化有一个显著的特点，那就是能够将事物给数字化和隐秘化，如钥匙、卡类传统身份管理物件，随身携带容易丢失，还有容易被他人获取并伪造。相比之下，指纹就在手上，不会脱离，而且永远不会改变，这就决定着指纹识别在安全领域、司法领域能发挥重要作用。同时，不高的技术门槛和方便好用的特性也让指纹识别越来越融入人们的生活。

在人体生物特征鉴别技术的迅速发展中，产生了指纹自动识别技术，相较于其他身份鉴定技术，指纹识别技术被更广泛应用的原因有以下几个方面：

1)指纹是独一无二的，人与人之间不存在着相同的指纹；