Palmprint authentication technique using fuzzy embedded radon transform and Collaborative Representation based Classifier

Abstract

Biometric authentication techniques have seen a resurgence in popularity and application recently as the need for secure data access increases. In this paper, we present a novel method for authentication of palmprints, which are known to serve as a robust biometric trait, yet has remained underutilized. We propose a modification of the radon transform incorporating fuzzy membership function which allows superior texture extraction of palmprint images. The application of Collaborative Representation Classifier is explained in this context, and a comparative study is presented in favor of the merits of our modification of the radon transform. The overall goal is to develop a new and efficient technique whose results can be translated to any database, allowing a variety of applications.