Adaptive Principal Components Extraction Algorithm and Its Applications in the Feature Extraction of Human Face

In this paper,the common grounds and differences between E.Oja's idea of statistical principal components analysis in recursion network and S.Kung's adaptive principal components extraction (APEX) algorithm in feature extraction are pointed out.The convergence of the algorithm is proved.Simulation results demonstrate the convergence and stability of the algorithm.It is analyzed that principal components number,subimage size and learning rate have effect on the algorithm.Therefore,it is expressed that the algorithm is a valid feature extraction method with less operation in face recognition.