Security technology – facial recognition

Abstract

In these recent years, face recognition become a very popular field of study and research subject in computer version. Both computer science researchers and the psychologists are showing a great interest in this subject because of the nature of the problem. Therefore more and more technologies and approaches are developed during the period. In this paper, an analytic approach is used to develop the system of face recognition. The database consists of almost thirty frontal-view face for testing. Firstly, the system is developed to find the human face boundary with the snake approach. Secondly, the system will able to find the 15 feature points on the head model. The feature points will consist of eyes position, nose position, mouth position and others facial feature position. Then these feature points will compare with the points which stored in the database. Finally, likely matched faces will be showed. The overall recognition rate is around 80 percent. As the system is designed to recognize just front-view faces, therefore faces form different viewing direction is not suitable for this face recognition system.