Face image recognition using PCA, 2DCT and neural network

Abstract

Automatic face recognition is a very important research area in computer science since it has been widely used in security systems. During the last two decade, the research area of face recognition has focused much attention from the scientific communities with the aim to provide highly intelligent human-machine interaction with high performance. This study proposes a system that encompasses a reduction of significant variable features using the principal components analysis on one hand, feature extraction using 2DCT on the other hand and then a combined method using both. Afterward each sample image is classified according to their pattern class using a Neural Network. The experimental results obtained shows an improvement in term of recognition rate when we combined the two methods.