Title

A Novel Facial Recognition Technique Employing Karhunen-Loeve Transform

Abstract

A Transform Domain Two-Directional Two-Dimensional Principal Component Analysis (TD/2D2DPCA) algorithm applied to facial recognition is presented. This algorithm has attractive properties with respect to storage and computational requirements while maintaining high recognition accuracy. The storage requirements are reduced by more than 95 percent compared to the spatial 2DPCA method and 25 percent compared to the TD2DPCA algorithm. The computational speed, compared to the spatial 2DPCA method, is reduced to a great deal. Experimental results obtained by applying the new algorithm to the ORL and Yale databases confirm these excellent characteristics.

Publication Date

12-1-2006

Publication Title

WMSCI 2006 - The 10th World Multi-Conference on Systemics, Cybernetics and Informatics, Jointly with the 12th International Conference on Information Systems Analysis and Synthesis, ISAS 2006 - Proc.

Volume

5

Number of Pages

217-220

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

Socpus ID

84867249026 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/84867249026

This document is currently not available here.

Share

COinS