Biometric Modality Benchmark Analysis

Abstract

Biometric modalities are templates or features generated from Biometric Traits like Retina, Fingerprint, Face, Palm Print and So on. There are several biometric techniques proposed over the years which meets high recognition rate and many of them are commercially available. For a new application, it becomes difficult to select a particular biometric trait or technique. Such selection should be based on comprehensive performance analysis of various biometric modalities. There are not many concrete proposals towards this direction of performance benchmarking for allowing the user to select a particular one. In this work we propose a unique methodology for analyzing the performance of various Biometric template generation techniques using several benchmarking parameters such as time invariance, Templates quality etc. Unlike normal text password based authentication problem, such variability‟s cannot be accepted as unlike text passwords, biometric templates cannot be changed. Even slightest of the variability in the template changes the generated key, henceforth causing higher false rejection rate. Hence in this work we analyze the most accepted biometric features and techniques for template generation and propose most invariable technique in terms of data acquisition invariability. This work analyzes Iris, Face, Fingerprint and Palm prints for analysis of the biometric template generation and key generation form the templates.