Machine Vision for an Intelligent Tutor

This paper presents a facial expression analysis system that performs recognition and emotional classification of human facial expression from a full-face image. The system consists of four main components. The first component performs face detection in an unstructured image using Artificial Neural Network. The second component is face recognition using Principal Component Analysis. The third component is facial feature extraction, which converts pixel base facial information into high-level geometry information. The fourth component is a fuzzy facial expression classifier. People’s facial features are analyzed logically and a decision unit is used to classify the expression. The system was implemented using extensions of existing algorithms and a new fuzzy approach to the classification of facial expressions.