Biometric Authentication via Facial Recognition

Security has become a vital need throughout the globe. An underdeveloped country Pakistan wants a reliable security system to prevent from the increasing threats of terrorism. A tool of artificial intelligence-biometric authentication can be a solution for such a problem. By implementing authentication system based on facial recognition in highly crucial areas of the country, the terror attacks can be controlled notably. Supervised learning can support this methodology. In this paper we used Principal Component Analysis (PCA) as a feature descriptor and for dimension reduction, K-Nearest Neighbor (KNN) as a main classification technique for the facial identification. The results of this combination applied on MIT face database are reported. We got up to 98.66% accuracy in detection rate using this combination.