Abstract
This paper presents an approach to recognize human facial expressions for human-robot interaction. For this, the facial features, especially eyes and lip are extracted and approximated using Bezier curves representing the relationship between the motion of features and changes of expressions. For face detection, color segmentation based on the novel idea of fuzzy classification has been employed that manipulates ambiguity in colors. Experimental results demonstrate that this technique can robustly classify skin region and non-skin region. In order to decide whether the skin region is face or not, largest connectivity analysis has been employed. This method can recognize the facial expression category, as well as the degree of facial expression change. Finally, the system has been implemented by issuing facial expression commands to a manipulator robot.