Face authentication using logic fusion of Color with EFM Feature Extraction

Abstract

In this paper, we investigate the use of colour information to face authentication systems. In order to improve the performance of these systems many colour components have been used. The results in different colorimetric components are combined by using a logic fusion for classification with different operators. For the extraction of feature vectors we have applied the Enhanced Fisher linear discriminant Model (EFM) which is presented as an alternative features extraction algorithm to Principal Component Analysis (PCA) which is widely used in automatic face recognition. We calculate the error rates in the two sets of data validation and data test according to the Lausane protocol (XM2VTS). The results obtained show that the use of colour improves the performance of authentication by (3%) compared to greyscale system. This system can be employed in high security.