Feature Extraction Technique of PCA for Face Recognition With Accuracy Enhancement

Abstract

Face recognition is a very complex task in the area of image processing and computer vision. This becomes important because it applies on many real life applications like Security, identification, crowd surveillance, Video surveillance etc. This paper is proposed the PCA based Face recognition. PCA is a holistic based Statistical method which is used to extract the feature from face image and to decrease the large dimensionality of the data to the smaller dimensionality of feature space, then classification is done using Euclidian distance classifier to recognize the face. The proposed method is worked on Yale Database and evaluate under varying conditions like Illumination variant for Center, Left and Right, Different Facial Expression like Happy, Sad, Normal, Wink , Surprised and Sleepy , Wearing spectacles and without it for Frontal Face View. The proposed work demonstrates the recognition rate for given Dataset.