“…We are researching on technology that extract human feature data from input picture image, about face, body and motion, and synthesize human image with computer graphics [3]. These technology is called "Model based recognition and synthesis", and i t is used for future TV telephone with very low bit rate$] and for multimedia communication [2], In this paper, we reports a system that can recognize facial image automatically, and can synthesize facial simulation image such as facial expression.…”