Facial Expression Recognition Using Facial Expression Intensity Characteristics of Thermal Image

摘要

To develop a robot that understands human feeling, we propose a method for recognizing facial expressions. A video was analyzed by thermal image processing and the feature parameter of facial expression, which was extracted in the area of the mouth and jaw by a two-dimensional discrete cosine transform. The facial expression intensity, defined as the norm of the difference vector between the feature vector of the neutral facial expression and that of the observed one, was measured. The feature vector made by facial expression intensity and time at utterance was used for recognizing facial expression.