Using Linear Regression Analysis for Face Recognition Based on PCA and LDA
Gang Xu, Shengli Zhang, Yunyun Liang · 2009
Aiming at pose-invariant face recognition, this paper proposes a method of generating virtual frontal view from a non-frontal face image, we use the virtual frontal face image instead of the input non-frontal image as the test image with the exertion of principal component analysis (PCA) and linear discriminant analysis (LDA) to finally achieve face recognition procedure. In this paper, this method is defined as "G+PCA+LDA". The experiment results show that the synthesis of virtual face generation and PCA+LDA face recognition algorithm has been proved to be a better way solving pose problem in face recognition than the classical algorithm such as principal component analysis and linear discriminant analysis.