A Fuzzy Type-1 Facial Recognition System

This paper presents a type-1 fuzzy logic controller used for face recognition. The process of identifying a face of an individual can be summed up in three phases; face detection, feature extraction, and face recognition. In this system, we are only concerned with the last phase; i.e. the face recognition of the human being. In this paper, we propose a fuzzy type-1 inference system that solely handles the facial recognition process. The reason we decided to tackle the facial recognition problem using fuzzy inference is (1) the use of fuzzy set theory in membership functions allows us to intuitively collect, classify and categorize our training data, (2) the fuzzy inference system (i.e. the if-then rules structure) gives us an intuitive reasoning that mimics the human way of thinking. We have tested our system and compared it with existing facial identification models, and it showed superiority in performance. This is because fuzzy logic is a powerful tool that is able to handle uncertainties existing in data; in our case the person’s facial image.