金库门禁指纹识别系统算法研究

发布时间：2018-03-24 05:22

本文选题：图像处理　切入点：质量评估　出处：《西安工程大学》2016年硕士论文

【摘要】：指纹身份验证方式具有便捷、唯一等可作为密码的天然特性。考勤记录、社保管理、门禁安全、车锁密码等指纹识别应用遍布各个领域。然而,由于计算机技术的各种弊端、传统指纹算法的安全漏洞等不利因素的影响,基于指纹身份安全认证的系统应用在金融、海关、军队等关系国家机密的、要求等级较高的、信息保密性较强的领域中,依然存在一些安全问题与应用问题。本课题针对特殊领域的指纹识别系统进行研究,对于军队、银行中的安全管理系统的稳定运行与快速发展影响深远。在了解当前指纹算法以及指纹应用的研究现状的基础上,对指纹识别算法学习研究和实验测试的数据进行深入的分析,找出其不足与缺陷,本文提出了克服其缺陷的算法,并结合金库门禁特殊的安全管理控制要求,最终设计一套金库门禁身份认证系统方案,提高了金库门管理的安全系数。主要研究内容包括:学习与研究当前的指纹识别技术方法的理论知识,对指纹识别应用做进一步分析;针对图像质量评估、增强等核心算法进行不断改进,具体体现在以下几个主要方面。首先,对指纹进行质量评估分类方法进行改进。传统方法主要是基于单一指标从图像的某一方面来判定其质量等级,有一定的局限性,不全面性。针对这点,本文提出了全局信息与局部信息相互度量的多指标质量评估方法。该算法从不同的角度,采用不同的测量指标衡量图像的不同特性,综合评估图像质量,提高了图像质量评估的准确度。其次,在计算图像像素频率时,本文给出了改进的边缘块频率的方法,实验证明该算法不仅能够准确地计算出纹理边界处的频率,而且改善了Gabor滤波对边缘区域的增强效果,对后续特征点的完整提取奠定基石。再次,在指纹特征提取及匹配阶段,文中比较了不同的算法,根据本课题的应用场景,选用基于细化模板匹配法,准确快速地提取了图像特征点,并采用基于细节点匹配方法来判断两幅图像是否为同一指纹。最后,结合金库门禁安全级数较高的要求,将质量评估、增强匹配、报警系统等各个加强安全级数的模块联合,设计了金库门禁指纹识别系统。测试表明,该系统不但运行稳定,而且提高了安全系数,加强了信息保密性。通过验证,本文的金库门禁指纹识别系统算法准确度高,稳定度高,可以对指纹图像进行准确的质量分类、识别匹配,实现了工作人员身份的准确验证,为金库门安全管理提供有力的保障。
[Abstract]:Fingerprint authentication method is convenient, unique and can be used as the natural characteristic of password. Attendance record, social security management, entrance security, car lock password and other fingerprint identification are widely used in various fields. However, due to the shortcomings of computer technology, Because of the negative factors such as the security loophole of traditional fingerprint algorithm, the system based on fingerprint identity security authentication is applied in the fields such as finance, customs, military and other fields which are related to state secrets, require higher level, and have strong information confidentiality. There are still some security problems and application problems. The stable operation and rapid development of the security management system in the bank have far-reaching influence. On the basis of understanding the current situation of fingerprint algorithm and fingerprint application, the study of fingerprint identification algorithm and the data of experimental test are deeply analyzed. To find out its shortcomings and defects, this paper puts forward an algorithm to overcome the defects, and finally designs a set of access guard identity authentication system scheme, combining with the special security management and control requirements of the vault access control. The safety factor of treasury door management is improved. The main research contents include: studying and studying the theoretical knowledge of current fingerprint identification technology, further analyzing the application of fingerprint identification, and aiming at image quality evaluation, The core algorithms, such as enhancement, are continuously improved in the following main aspects. First of all, The traditional method is mainly based on a single index to judge the quality grade of the fingerprint from one aspect of the image. It has some limitations and is not comprehensive. In this paper, a multi-index quality evaluation method for measuring global and local information is proposed. The algorithm uses different measurement indexes to measure the different characteristics of the image from different angles and synthetically evaluates the image quality. The accuracy of image quality evaluation is improved. Secondly, when calculating the pixel frequency of the image, this paper presents an improved method of edge block frequency, which is proved by experiments that the algorithm can not only accurately calculate the frequency at the edge of the texture. Moreover, the enhancement effect of Gabor filter on edge region is improved, which lays the foundation for the complete extraction of subsequent feature points. Thirdly, in the phase of fingerprint feature extraction and matching, different algorithms are compared, according to the application scene of this subject, Based on the thinning template matching method, the feature points of the image are extracted accurately and quickly, and the minutia-based matching method is used to judge whether the two images are the same fingerprint. By combining the modules of quality assessment, enhanced matching, alarm system and so on, the paper designs a fingerprint identification system for access control in vaults. The test results show that the system not only runs stably, but also improves the safety coefficient. Through the verification, the algorithm of this paper has high accuracy and high stability, it can classify the fingerprint image accurately, identify and match the fingerprint image, and realize the accurate verification of the identity of the staff. For the Treasury door security management to provide effective protection.
【学位授予单位】：西安工程大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41

【相似文献】