The New Face Recognition Technique With the use of PCA and LDA

Image recognition using various image classifiers is an active research area. In this paper we will describe a new face recognition method based on PCA (Principal Component Analysis) and LDA (Linear Discriminate Analysis). The new method consists of two steps: first we project the face image from the original vector space into a face subspace by using PCA technique and finally we use LDA to obtain a linear classifier. The idea of combination of two methods proved enough good improvement in the generalization capability of LDA where only few samples per class were available. BY using ORL dataset we observed a significant improvement compared with original images were fed directly to the LDA classifier. The newly proposed hybrid classifier has provided a useful framework means for general image recognition tasks as well.