An Improved Framework for Human Face Recognition

In recent years considerable progress has been made by the researchers in the field of pattern recognition in general and face recognition in particular. Computers can now outperform human brain in face recognition and verification tasks. While most of the methods related to face recognition perform well under specific conditions, some show anomalous behavior when the degree of accuracy is concerned. In this paper, we have divided the face recognition task into three sub-parts as Segmentation, Feature Extraction, and Classification. Information from face image is extracted and modelled using Eigenvectors. The weights calculated from Eigenvectors are classified by the statistical classifier using distance metric specification. The system is capable of recognition to an accuracy of 96%, having a standard deviation of 0.662 for facial expression variations.