文档介绍：湖南科技大学
毕业设计( 论文)
题目
基于神经网络的指纹识别系统研究
作者
学院
专业
学号
指导教师
年月日
摘要
随着信息技术的和网络技术的发展,信息安全越来越引起人们的重视。为了保护,自身的信息、财产、以及资料安全,很多场合和设备都需要对来访者和使用者进行身份验证。指纹识别由于具有唯一性、可靠性,基于指纹识别的技术的身份识别系统以其独特的技术优势和成本优势广泛的应用于个领域。
目前指纹匹配系统的主要问题存于这几个方面:在系统采集的指纹图像易受噪声和皮肤弹性等因素影响;特征匹配的效率较低;容易受伪特征点的影响;对模糊和有一定位移、旋转的图像无法准确识别等。本文将神经网络技术与模式识别技术相结合,提出一种对解决指纹识别技术中的问题切实有效的解决办法。概括起来,本文主要对指纹识别系统的以下几个方面进行了深入研究。
本文给出了用matlab对指纹图像的处理算法以及处理结果。主要使用了,图像的频域增强,自适应阀值二值化和指纹的细化。实验显示,matlab实现的处理结果比较理想,满足识别的应用性。
在指纹的特征点提取技术研究中,本文采用了全局的一个指纹走向特征和局部的细节特征点作为神经网络识别的特征,从而达到准确识别指纹图像的目的。
在指纹的识别方法中,本文使用了一种基于BP神经网络的指纹识别算法,使用神经网络的识别,能有效的提高指纹识别的鲁棒性。
关键词:BP神经网络;指纹图像预处理;指纹特征提取;伪指纹特征处理
Abstract
With the development of information technology work technology, more and more people pay attention to information security. In order to protect their own information, property, as well as data security, on many occasions and equipment needs of visitors and user authentication. Fingerprint recognition due to unique, reliability, identification systems based on fingerprint recognition technology with its unique technical advantages and cost advantages of a wide range of applied field.
Fingerprint matching system stored in these aspects: the fingerprint image acquisition by the system is susceptible to noise and skin elasticity and other factors; less efficient feature matching; vulnerable to the impact of false feature points; blur and a certain displacement, the rotation of the image can not accurately identify. work and pattern recognition technology is presented bining an effective solution to solve the problem in fingerprint recognition technology. To sum up, this article is mainly on the following aspects of the fingerprint identification system in-depth study.
In this paper, using matlab fingerprint image processing algorithms and processing results. Mainly spent, in the frequency domain of the image enhancement, adaptive threshold binarization and thinning of the fingerprint. The experiments showed that the Matlab achieve re