Abstract

The face as one part of the human body has been widely used as a biometric-based security system. This research is a process of designing and making of the identification system which is able to detect of human attended. This system implements viola-jones, eigenface, and Euclidean distance methods. Viola-jones method is employed as face detection. Eigenface method is employed to reduce the face image via a projection. The Euclidean distance is implemented as a classification method. The accuracy of this system against 30 register subjects and 15 unregistered subjects is 84%. This accuracy is obtained after the addition of face normalization and adaptive training model on the system. Therefore, this research concluded that viola-jones method is very good as face detection, however, need to be added face correction process. Meanwhile, eigenface and Euclidean distance methods provide good results to recognize the face when many training faces are given.

Concepts :

Access to Document

Citations by Year