Analysis and recognition of the facial expressions play an important role in both the social society and the affective computing in the field of the computer science. There are three primary methods for the analysis of expression motive features: methods based on the facial geometrical structure features, on the definition of the expression space based on the eigen-face, and on the motion pattern matching. This paper extracts the feature regions of the expressions based on the facial physics-muscle model and evaluats the optical flow of the expression image sequences. The eigen-flow vectors can be calculated to constitute the eigen-sequences, and therefore, the expressions can be analyzed. The recognition system is implemented as an agent in the multi-perception machine and it is used as part of the video input for understanding the human body languages.