基于嵌入式系统的指纹识别算法的研究

发布时间：2018-01-20 03:06

本文关键词： 指纹识别指纹分割滤波增强嵌入式特征提取　出处：《武汉科技大学》2012年硕士论文　论文类型：学位论文

【摘要】：随着社会的飞速发展，对个人身份识别的需求无处不在，并且与日俱增。传统的身份识别方法的局限性和弊端表现得越来越明显。悄然兴起的生物特征识别技术，可以从本质上解决传统了身份识别技术的不足，因此正逐渐被认可。指纹识别技术发展至今已经成为生物特征识别应用最为可靠，方便的一种类型。传统的基于PC机的指纹识别系统存在着体积大，不易携带，实时性差等特点。这些使得基于嵌入式系统的指纹识别系统成为正在研究的热点。随着移动通信蓬勃发展和便携式设备的推陈出新，，指纹识别技术在今后将被更多的应用于便携式设备。在这一背景下，基于嵌入式的指纹识别技术的研究无论对于让指纹识别从依赖于PC机向脱机使用发展以提高系统灵活性，实用性，性价比，还是对于拓展嵌入式系统的应用领域，都具有重大意义。本文中，首先对指纹识别技术的特点、应用和市场概况做了研究。讨论了一系列嵌入式指纹开发涉及到的技术问题，给出了嵌入式系统一般开发步骤。对于指纹采集仪设备理论和参数等做了研究，并对几种类型的指纹采集仪做了对比。然后，对基于ARM嵌入式的指纹识别系统进行了研究。从系统硬件和软件两个方面对嵌入式指纹识别系统进行了设计，分析了设计开发过程中可能遇到的问题，给出了硬件设计框图及设计思路，并对软件算法流程作了分析。最后，着重研究了适用于嵌入式系统的指纹识别算法。提出了一种改进的指纹图像分割方法；对基于Gabor滤波器的指纹图像增强算法进行了改进；提出了改进的细化方法对指纹图像进行细化处理。通过针对指纹识别系统算法流程中的每一步骤分析，对现有的一些算法进行简化和创新，提出自己的改进算法，最后得出一整套适合于基于ARM的嵌入式系统上运行指纹识别算法。并将所有算法在PC机上编程实现。
[Abstract]:With the rapid development of society, the need for personal identification is everywhere. The limitations and drawbacks of traditional identification methods are becoming more and more obvious. The emerging biometric identification technology can solve the shortcomings of traditional identity recognition technology in essence. Fingerprint identification technology has become the most reliable and convenient type of biometrics application. Traditional PC-based fingerprint identification system has a large volume. These characteristics make the fingerprint identification system based on embedded system become a hot topic. With the rapid development of mobile communication and the innovation of portable devices. Fingerprint identification technology will be more used in portable devices in the future. In this context. The research of fingerprint identification technology based on embedded system can improve the flexibility, practicability and cost performance of the system by developing fingerprint identification from PC to offline. Or to expand the application of embedded systems, are of great significance. In this paper, first of all, the characteristics, application and market situation of fingerprint identification technology are studied, and a series of technical problems involved in the development of embedded fingerprint are discussed. The general development steps of the embedded system are given. The equipment theory and parameters of the fingerprint collector are studied, and several types of fingerprint acquisition instruments are compared. The fingerprint identification system based on ARM is studied. The embedded fingerprint identification system is designed from two aspects of system hardware and software, and the problems that may be encountered in the process of design and development are analyzed. The hardware design block diagram and design idea are given, and the software algorithm flow is analyzed. Finally, the fingerprint recognition algorithm suitable for embedded system is studied emphatically, and an improved fingerprint image segmentation method is proposed. The fingerprint image enhancement algorithm based on Gabor filter is improved. An improved thinning method is proposed to refine the fingerprint image. By analyzing each step of the fingerprint identification system algorithm, some existing algorithms are simplified and innovated. An improved algorithm is proposed and a set of fingerprint recognition algorithms which are suitable for the embedded system based on ARM are obtained and all the algorithms are programmed on PC.
【学位授予单位】：武汉科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP391.41;TP368.1

【参考文献】