CORC  > 自动化研究所  > 中国科学院自动化研究所  > 毕业生  > 博士学位论文
题名指纹匹配算法的研究与应用
作者张堂辉
学位类别工学博士
答辩日期2003-05-01
授予单位中国科学院研究生院
授予地点中国科学院自动化研究所
导师田捷
关键词指纹识别 相似度直方图 细节匹配 纹理匹配 生物特征识别 fingerprint identification similarity histogram minutia matching texture matching biometrics
其他题名Fingerprint Matching Algorithm and Application
学位专业模式识别与智能系统
中文摘要计算机和网络技术的发展对身份验证技术提出了更高的要求。传统的ID+ 密码存在着密码容易被窃取和记忆困难的缺点,因此人们把目光转向了生物特 征识别技术,而指纹是最可靠的生物特征识别技术之一。虽然在指纹识别方面 已经有大量的研究工作,但是指纹识别领域仍然存在着很多挑战性的问题。例 如指纹特征的提取,低质量指纹图像的增强、非线性形变指纹的匹配等。在本 论文中,重点针对指纹的匹配问题进行了深入的研究,并且获得了一些有价值 的研究成果。本文的主要贡献可以概括如下: (1)提出了一种基于统计相似度的快速指纹匹配算法-SHA算法。该算法通 过建立相似度直方图函数来计算指纹校准参数,并且将指纹纹线信息引入了指 纹匹配中,有效地提高了指纹识别算法的性能。 (2)提出了一种改进的指纹混合匹配算法。该算法在SHA细节匹配算法的 基础上,融合了指纹方向纹理和Gabor滤波纹理进行混合匹配。该算法在纹理 特征的提取上采用了方形网格的分块法,使得计算更加简洁和方便。FVC的实 验结果证明了该算法的有效性。 (3)设计并实现了一个用于指纹识别研究专用的指纹图像处理系统。这个图 像处理系统有很丰富的图像显示功能,支持指纹算法的动态加载、FVC和BioAPI 标准。该系统的开发为我们指纹算法的研究打下了一个良好的基础。 (4)提出并实现了一种DSP嵌入式指纹识别系统的核心算法。我们重点研 究了在嵌入式系统上进行指纹算法化简的问题,并提出了算法串行化处理的概 念。也对算法实现时碰到的混合编程,算法调度等关键技术问题进行了研究和 探讨。
英文摘要With the rapid progress of computer science, an accurate and reliable personal identification technology is needed. The conventional methods (PIN+password) have a number of disadvantages: It can be stolen, forgotten or guessed by impostors. Biometric is adopted to solve these problems. Fingerprint-based authentication systems have received the most attention because it is one of the most reliable biometric technology. A lot of research work has been reported in the last decades. However, there are still challenging problems, such as reliable feature extraction, enhancement of poor quality images, fingerprint matching of highly distorted fingerprint images. In this thesis, we focus on the fingerprint matching problems. The main contribution of this thesis includes: (1) A fast fingerprint matching algorithm called similarity histogram approach (SHA) is proposed. The accurate alignment parameters are obtained by similarity histogram function. We also introduce the ridge information in the global matching stage. Experimental results on FVC databases show it improved the fingerprint performance effectively. (2) We present a hybrid approach to fingerprint matching that combines SHA algorithm and a texture-based matching algorithm. Texture features are extracted from rectangular tessellation of directional image and gabor-filtering images. The experimental results show EER is reduced by 1.8%. (3) The design and implementation of our fingerprint image processing system is introduced. It provides abundance functions, such as image display, dynamic loading, FVC benchmark testing and BioAPI interface. The system is helpful for our research. (4) We design and implement an algorithm on a DSP embedded system. We focus on "serialized algorithm" in the process of algorithm simplification. The solutions of mix-language programming and code scheduling are also present in this paper.
语种中文
其他标识符749
内容类型学位论文
源URL[http://ir.ia.ac.cn/handle/173211/5763]  
专题毕业生_博士学位论文
推荐引用方式
GB/T 7714
张堂辉. 指纹匹配算法的研究与应用[D]. 中国科学院自动化研究所. 中国科学院研究生院. 2003.
个性服务
查看访问统计
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。


©版权所有 ©2017 CSpace - Powered by CSpace