节点文献

基于DSP的指纹处理算法研究与实现

Research and Realization of Fingerprint Disposal Algorithm Based on DSP

【作者】 陈庆接

【导师】 李见为;

【作者基本信息】 重庆大学 , 仪器科学与技术, 2005, 硕士

【摘要】 传统的安全系统主要采用基于信物或口令的方式,随着社会的发展,这种系统显得越来越脆弱。为了应对这些挑战,人们把目光转向了生物识别技术,希望能借助人体的生理特征或行为动作来提高身份识别的精度。所谓指纹就是在人们的手指表面上的凸凹不平的各种各样的纹路。现代统计学已经表明,指纹具有稳定性和唯一性,也就是说,每个人的指纹终身不变而且与其他人的指纹不同。最近,随着计算机技术和嵌入式系统技术的发展,自动指纹识别技术取得了很快的发展,指纹图像处理技术是自动指纹识别系统的关键技术之一,本论文重点介绍的就是这种指纹图像处理技术。作者在收集和分析了近年来大量的国内外的关于指纹图像处理技术的学术文献基础上,分析了指纹图像的特性,提出了对指纹图像的一些合理的假设,并从这些假设出发,提出了一套适用于电容式指纹传感器采集的指纹图像的处理算法,这些算法包括指纹图像的质量分析、指纹图像的分割、指纹图像滤波、二值指纹图像的细化、特征点的提取、特征点的可靠性分析、指纹图像的点模式匹配等步骤。指纹匹配算法的研究是自动指纹识别的一个十分重要的研究内容。针对基准点定位困难的问题,本文提出了一种根据特征点集中三个近邻点所组成的子集之间的相互关系来确定基准点以及变换参数的算法。作者工作的最终目标是实现一个嵌入式DSP 指纹识别系统,嵌入式指纹识别系统是一个相对独立的完整系统,它不需要连接其他设备或计算机就可以独立完成其设计的功能。由于采集系统硬件本身的原因,所采集的指纹并不是理想的,通过一种有效的指纹预处理算法,提高系统对采集的指纹的适应性是十分重要。由于PC 端的程序设计与DSP 嵌入式系统的编程存在着很大的差异,解决算法由PC 向DSP 系统的移植同样很重要,本论文的工作就是介绍了一种指纹识别算法,以及如何实现算法由PC 端到嵌入式DSP 系统的移植。

【Abstract】 Traditional security system mainly is based on token or password. With the development of society, this system has become more frail. To cope with this challenge, man has turned look of biometrics and hoped to enhance identity verification by using his physical character and behavior. A fingerprint is the pattern of ridges and valleys on the surface of a fingertip. Modern statistics has proved that fingerprint is unique and steady, that means, one’s fingerprint is different from all others and will not change all his life. Recently, as the development of micro –computer technique and embed developing system technique, automated fingerprint identification system (AFIS) has made a great progress. This article is meanly stress on the fingerprint image processing technique that is one of the key techniques of automated fingerprint identification system. This article analyses the characteristic of fingerprint image and brings forward some reasonable supposes which related to the fingerprint image after I collected and analyzed many articles of others that is related to the fingerprint image processing and has published in the resent years. Based on these characteristic and supposes, I developed a set of fingerprint image processing arithmetic that are adapt to the fingerprint image gathered by capacitive solid-state device. These include several steps. They are analyzing quality, segmentation, filtering noise, thinning, refining the minutiae, analyzing reliability and matching of fingerprint image. Fingerprint matching is an important step of AFIS. The paper proposes a novel algorithm, which acquires a pair of reference minutiae and transform parameters for fingerprint matching according to the relationship between two subsets that consist of three neighboring minutiae derived from the two minutiae sets. The target of author is to make an embed automated fingerprint identification system, this means this system can finish the designed function without any other equipment or computer. For the reasons of sampling system, the fingerprint image can’t meet our needs for identification. Using the pretreatment to the fingerprint image to enhance the adaptability of AFIS to the different condition is effective way. At the same time, there is great difference between programming on PC and embed system. So how to resolve the problem of transplant the arithmetic from PC to embedded system is anther question.

【关键词】 指纹图像图像处理匹配DSPC5410
【Key words】 FingerprintImage processingmatchingDSPC5410
  • 【网络出版投稿人】 重庆大学
  • 【网络出版年期】2005年 08期
  • 【分类号】TP391.41
  • 【被引频次】10
  • 【下载频次】593
节点文献中: