IMPLEMENTATION OF VIEWING ANGLES FOR FACE RECOGNITION IN CONSTRAINT ENVIRONMENT

Abstract

The objective of the project “IMPLEMENTATION OF VIEWING ANGLES FOR FACE RECOGNITION IN CONSTRAINT ENVIRONMENT” is to implement face recognition algorithms on the images stored within database. The images used in this project have been acquired from the standard “AT&T face database”. This database is an open source database, which stores grey scale pictures of some subjects from different viewing angles. The database, stores images of forty subjects (humans), each clicked from ten different viewing angles. The project implements PCA (Principal Components Analysis) using Eigen Faces and K Nearest Neighbour Algorithm Principal Components Analysis reduces dimensionality of complex training set and Eigen analysis to convert the data sets into Eigen vectors. The face vectors then normalized, and Eigen face vectors are calculated. uses The sections of this report further elaborate the algorithms and techniques deployed in this project.