Effective Face Recognition System Using Hybrid Principal Component Analysis For Real-Time Application

Abstract

:- In Principal component Analysis the facial Features are in form of Eigen-values. In Linear Discriminant Analysis (LDA) the Facial features are in form of Fisher Discriminants. The major disadvantage of the existing system is that minute Discriminants are impossible due to the application of binarylogic. In this paper a Hybrid Principal Component Analysis (HPCA) face-verification method is proposed for the high-resolution face authentication task with the eigenapproach. The Wavelet is a Powerful tool to analyze and process image signals. It is mainly used for dimensionality reduction. The facial expressions with low dimensional subspace can be obtained through the Fishers linear discriminant on fuzzy method.  Variations in light and variations in facial expressions will not affect the recognition. The proposed algorithm helps to future extraction based on local and global trained features and increases the accuracy Upto 3% percentage.it is MATrix  LABoratory tool (MATLAB) the experimental results shows recognition rate increases compare to other technique.