Resumo

SILVA MATA, Francisco José et al. Biometric recognition of faces through the analysis of functional data of their three-dimensional models. Rev cuba cienc informat [online]. 2017, vol.11, n.1, pp. 1-14. ISSN 2227-1899.

ABSTRACT The representation of face data recorded in three dimensions with the aim of identifying people, while solving the problems of pose and illumination that affect the performance of two-dimensional systems, faces the problem of high data dimension, which causes classification errors because of the so-called "curse of dimensionality". For this reason, the reduction of the dimension maintaining high recognition rates is a priority objective of the methods of representation of the biometric objects. As a result of this research, we propose the representation of three-dimensional information through functions, achieving a compact representation that provides additional advantages such as representing the correlation between neighboring values, and a notable simplification of computational computation. The results obtained demonstrate the validity of the functional analysis as a way of recognition of faces in three dimensions complete and with lost parts.

Palavras-chave : Functional data analysis; biometric recognition; dimension reduction.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )