Adaptive Modified PCA for Face Recognition.

摘要

In many real-world applications such as face recognition and mobile robotics, we need to use an adaptive version of feature extraction techniques. In this paper, we introduce an adaptive face recognition system based on PCA algorithm. We combine Sanger’s adaptive algorithm for computation of effective eigenvectors with QR decomposition algorithm where used to estimate associated eigenvalues. By normalizing extracted feature vectors, we construct a new more effective feature subspace and used it for on-line face recognition. Experimental results on Yale face data base demonstrated the effectiveness of proposed system in real-time face recognition applications. 1.