Self-adaptive biometric classifier working on the reduced dataset

Abstract

The paper presents a method of object recognition by means of a reduced data set. These data are specially prepared. The proposed method was also compared with two other well-known data reduction techniques, Principal Component Analysis (PCA) and Singular Value Decomposition (SVD). Objects can mostly be described through many features but these features can have different discriminant powers. The Hotelling's statistical method, allows determining the best discriminatory features and similarity measures which can be simultaneously selected. © 2014 Springer International Publishing.