Functional Data Analysis as an Alternative for the Automatic Biometric Image Recognition: Iris Application

Abstract

Functional data analysis has been a novel option for representing images, since the continuous nature of images is preserved. Image representation using functional data provides significant advantages, being the appreciable reduction of the dimensionality one of the most significant. This paper gives a detailed description of the entire imaging process using the proposed approach. As an example, the representation of iris images through functional data for recognition tasks was used. The paper presents experiments and results of applying this approach to the recognition of iris images, demonstrating its effectiveness.