A Biometric Security Model with Co-Occurrence Matrices for Palmprint features

Abstract

This paper presents a biometric authentication model using the palmprint, in which the data processing is done using textural statistical features and co-occurrence matrices. Several features sets are extracted and then combined with a functional fusion rule, avoiding the commonly used concatenation feature-level fusion. The target application has several security levels in order to accurately identify the most authorized person. The designed system operates in identification mode because the goal is to guess the true identity of a person using the biometric credential without the user-name.