Biometric Authentication: An Hybrid Face Recognition System Model

A biometric authentication system based on face recognition is implemented and described in this work. The proposed algorithm is a combination of Principal Component Analysis (PCA) and a Feature based technique which is based on the two modalities used in face recognition i.e. The Holistic approach and Feature-based approach. PCA which was used as our mainstay algorithm, based on information theory concepts, seeks a computational model that best describes a face, by extracting the most relevant information contained in that face. With the feature based, we localized and measured some key features of the face such as the width and height of the face and the head midpoint. Based on this measurement, we generated a weight also for each image which we then combine with the weights computed from the eigenface approach. Recognition and authentication is performed by projecting a probe to the proposed system. Olivetti Research Lab (ORL) face database along with the face database we constructed (Colface face database) from images taken of fifty subjects of pure black population at varying pose and light intensities were used to test the system. A good and considerable result was achieved