Study of automated face recognition system for office door access control application

The security currently become a very important issue in public or private institutions in which various security systems have been proposed and developed for some crucial processes such as person identifications, verification or recognition especially for building access control, suspect identifications by the police, driver licenses and many others. Face recognitions have been an active area of research with numerous applications since late 1980s and become one of the important elements in security system development. This paper focuses on the study and development on an automated face recognition system with the potential application for office door access control. The technique of eigenfaces based on the principle component analysis (PCA) and artificial neural networks have been applied into the system. The study includes the analysis of the influences of three main factors of face recognition namely illumination, distance and subject´s head orientation on the developed face recognition system purposely built for office door access control. The experimental results have shown that the developed system has achieved good performance of face recognition rate of 80% at the distance of camera and subject between 40 cm to 60 cm and the subject´s orientation head angle must be within the range of-20 to +20 degrees.