Template security analysis of multimodal biometric frameworks based on fingerprint and hand geometry
Template security analysis of multimodal biometric frameworks based on fingerprint and hand geometry
复制标题
基于指纹和手部几何的多模态生物识别框架模板安全分析
DOI：
10.1016/j.pisc.2016.06.065
复制
发表时间：
2016
影响因子：
--
通讯作者：
Anubhuti
中科院分区：
文献类型：
--
作者：
A. Selwal;S. Gupta;Surender;Anubhuti
生物识别系统是用于在现代计算的各种应用中提供身份验证的自动工具。在这项工作中，提出了基于指纹和手几何模态的多模态生物识别系统的三种不同设计框架。还提供了一种分析来诊断所提出的系统中各种类型的模板安全问题。所有设计都应用了具有五个决策参数的模糊层次分析法（FAHP），并且发现框架1在模板数据安全性、模板融合和计算效率方面更好。值得注意的是，模板数据在存储到数据库之前的安全性是一项具有挑战性的任务。一个重要的观察是，可以在特征融合级别保护模板，并且可以使用索引技术来提高安全模板的大小。
Biometric systems are automatic tools used to provide authentication during various applications of modern computing. In this work, three different design frameworks for multimodal biometric systems based on fingerprint and hand geometry modalities are proposed. An analysis is also presented to diagnose various types of template security issues in the proposed system. Fuzzy analytic hierarchy process (FAHP) is applied with five decision parameters on all the designs and framework 1 is found to be better in terms of template data security, templates fusion and computational efficiency. It is noticed that template data security before storage in database is a challenging task. An important observation is that a template may be secured at feature fusion level and an indexing technique may be used to improve the size of secured templates.