指纹图像识别系统的研究与实现
2022-11-22 10:04:30
论文总字数:20835字
摘 要
生物识别技术,顾名思义,是指利用人类的固有生物特征,如笔迹、虹膜、掌纹等实现身份鉴别的技术,将生物特征转化成数字图像等传入计算机,结合其他各种高科技手段,实现身份认证。生物识别技术由生物学衍生而来,根据国际生物小组的统计结果显示,生物识别技术已经获得市场的喜爱,许多公司投入大量资金研发生物识别设备。其中,指纹识别技术不仅仅可以避免冒领或身份作家导致的财物丢失等经济损失,还避免了传统身份认证时的冗杂事务。指纹识别技术凭借其迅捷的识别速度、高度的稳定性和安全性以及可持续的发展性和指纹的可再生性,在市场中获得广泛推广。
本文利用Matlab搭建了一个指纹图像识别系统,实现了指纹图像的获取与匹配。首先,本文先详细介绍了研究指纹图像识别的原因及意义,简单介绍了指纹识别的发展史,并结合指纹识别系统的目前研究进度预测了其未来的发展前景。其次,本文简单介绍了目前图像处理应用最广的Python与Matlab各自的优缺点,并通过对比,选择更适合本系统的Matlab作为基础。接下来,本文研究了指纹图像识别的原理,从图像预处理、特征提取和特征匹配三个方面介绍了指纹图像识别的实现方法,并利用Matlab实现了仿真,且附有图像。其中,图像预处理包括指纹图像二值化及确定中心点、分割和灰度归一化以及Gabor非线性平滑滤波器的使用。指纹特征提取通过Gabor滤波器生成的八个方向的特征图提取特征值,通过计算待测图像与数据库图像的欧几里得距离实现指纹图像特征匹配,寻找出最相似的指纹。
关键词:指纹图像识别;二值化;灰度归一化;特征匹配;
Research and Implementation of Fingerprint Image Recognition System
Abstract
Biometrics technology, as its name implies, refers to the use of human's inherent biometrics, such as handwriting, iris, palmprint and other technologies to achieve identity identification, biometrics into digital images and other into the computer, combined with other high-tech means to achieve identity authentication. Biometrics technology is derived from biology. According to the statistics of the International Biological Group, biometrics technology has won the market's favorite, and many companies have invested a lot of money to develop biometrics equipment. Among them, the fingerprint recognition technology can not only avoid economic losses such as impersonation or loss of property caused by identity writers, but also avoid the tedious affairs of traditional identity authentication. The emergence of has solved the problem that the identification object is easy to be lost and forgotten in the traditional identification method, and at the same time overcomes the problem of economic loss caused by the identification object being stolen.
This paper uses Matlab to build a fingerprint image recognition system to achieve the acquisition and matching of fingerprint images. Firstly, this paper introduces the reason and necessity of fingerprint image recognition in detail, briefly introduces the history of the development of fingerprint recognition. Next, this article briefly introduces the most widely used image processing tools Python and Matlab advantages and disadvantages. Through comparison, Matlab, which is more suitable for this system, is selected as the basis. Then, this paper studies the aim of fingerprint image recognition, introduces the implementation method of it from three aspects: image preprocessing, feature extraction and feature matching, and uses Matlab to realize the simulation, with an image. Among them, image preprocessing includes fingerprint image binarization and determination of center point, segmentation and grayscale normalization, and the use of Gabor nonlinear smoothing filter. Fingerprint feature extraction. The eight direction feature maps generated by the Gabor filter extract feature values, and calculate the Euclidean distance between the image to be tested and the database image to achieve fingerprint image feature matching and find the most similar fingerprint.
Key words: Fingerprint Image Recognition; Fingerprint Image Binarization; Grayscale Normalization; Feature Matching
目 录
摘 要 I
Abstract II
第一章 绪 论 1
1.1 研究指纹图像识别技术的背景和意义 1
1.2指纹图像识别技术的国内外研究现状 1
1.3 本文的主要内容和行文结构 2
第二章 指纹图像处理方案与技术 3
2.1指纹图像识别方案设计论证 3
2.1.1图像处理平台的选取 3
2.1.2 Matlab基本简介 4
2.2 指纹图像识别系统基本结构与主界面 5
2.2.1系统基本结构 5
2.2.2系统主界面概况与介绍 6
2.3 本章小结 8
第三章 指纹图像的预处理 9
3.1 指纹图像的二值化 9
3.2 二值化后指纹图像中心点的确定 10
3.3 指纹图像的灰度归一化 11
3.4 Gabor非线性平滑的实现 12
3.5 本章小结 14
第四章 指纹特征提取与匹配 15
4.1 指纹特征提取 15
4.2 指纹特征匹配 15
4.3 本章小结 16
第五章 总结与展望 17
5.1 总结 17
5.2 展望 17
致 谢 19
参考文献 20
第一章 绪 论
1.1 研究指纹图像识别技术的背景和意义
信息时代,传统的身份鉴定方式显得手续繁琐且漏洞百出,常常被有心之人钻了空子,冒充。每年,我国因身份认证失败或冒充别人身份而导致的经济损失持续增长,拥有一种安全可靠又便捷稳定的身份鉴定方式显得越发重要,因此生物识别逐渐进入人们的日常生活。目前,已经有许多先进的生物识别技术用于身份鉴定,它们的共同点是安全性高,速度快,可靠性强。其中,指纹识别技术,以终生的不变性以及超高的稳定性获得了市场的喜爱,是几种生物识别技术中发展最为成熟的几个之一。由于不同的人指纹几乎完全不同,因此,指纹识别技术有很强的实用性,在商业活动、警务工作等需要身份鉴定的领域都有广泛的应用,在生物识别领域发展迅猛。随着电子科学技术的进步和算法的快速更新迭代,指纹图像识别技术是目前研究最为深入、发展最为成熟、最有应用前景的生物识别技术。
剩余内容已隐藏,请支付后下载全文,论文总字数:20835字