A Hierarchical Data Classification Model for Biometric Identification Systems

Abstract

The paper presents an innovative hierarchical classification model which should be applied for large-scale biometric identification systems, in order to improve their performance. The model is relying on a multi-level fusion approach, but completed with a feature-selection strategy. The practical application of the proposed model concerns networks security issues, especially for databases remote access, but also physical access control issues such as border or other secured sites.