Face recognition using PCA and geometric approach

Abstract

This paper presents a framework for the face recognition problem. In recent times face recognition had been paid ample attention from researchers but still remained confronting in real time applications due to the presence of noise. A wide range of face recognition techniques have been presented in the past few years which majorly fall under feature based or holistic based approaches. In our paper we use some ailments of geometric based approach which maps different fiducial points in the face and compares them, for effective recognition of faces and respective data retrieval. We use the principal component analysis algorithm in fusion with geometric approach for face recognition purpose. This paper demonstrates the power of our approach by using different experiments and vividly concentrates on the best similarity and proximity possible coupled with highest recognition rate.