Development of Pose Invariant Face Recognition Method Based on Pca and Artificial Neural Network.

Abstract

This paper considers human face for pose invariant face recognition as a biometric parameter. There are lots of methods proposed for the face recognition. Face recognition by Neural Network and PCA is common method used now days. But every time it is not possible that the situation or poses are same. Due to some reasons the conditions are change like wink position, blinking, left pose, right pose, etc. In such cases recognition goes difficult. This paper focused on pose invariant human face recognition system and as a result we focus on face recognition system using PCA with feed forward Neural (Multilayer) Networks for recognition of human face irrespective of pose of face in images. The Principal Components Analysis (PCA) is used as a feature extractor whereas Neural (Feed Forward) Network is used for classification. We refer Yale Face Database with 11 images for each subject in which is having poses like wink, open mouth, smile etc