Automatic Face Detection Using Color Based Segmentation and Face Recognition Using Eigen Face

Abstract

One of the most successful applications of image analysis and understanding are face recognition and face detection, which have recently received significant attention, especially during the past several years. At least two reasons account for this trend; the first is the wide range of commercial and law enforcement applications and the second is the availability of feasible technologies due to 30 years of research. The online services, such as E-Banks, Web-Mails, Social Site in which users are verified by their usernames and passwords, are increasingly exploited by Identity Theft procedures. Identity Theft is a fraud, in which someone pretends to be somebody else to steal money or to get other benefits. To overcome the problem of Identity Theft, an additional security layer is required and here Biometrics present potentially a good solution. In Physiological and Behavioral biometrics, users are verified and identified based on their keyboard and mouse activities. This thesis proposes a physiological based robust Face Detection and Face Recognition system based on HSV components and Eigen-face technique. Basically the purpose of Face Detection is to determine whether there is any face in an image, while Face Recognition involves confirming or denying the identity claimed by a person. The contributions of this thesis are the following: The first part of this thesis addresses a multi-views robust, improved segmentation algorithm for Face Detection employing HSV technique on color images with multiple-faces and skin tone. The goal of face detection is to identify all image regions which contain a face regardless of its three-dimensional position, orientation, and lighting conditions. Such a problem is challenging because faces are non-rigid and have a high degree of variability in size, shape, color, and texture. In the second part, a novel generative approach for Face Recognition, based on an Eigen-face description of the face is discussed. The result of this work is an algorithm that is fast & requires a simple training procedure and is highly efficient in bad lighting condition. After analyzing both algorithms and identifying their limitations, we conclude with several promising directions for future research.

ACKNOWLEDGEMENT