B\'ezier splines are widely available in various systems with the curves and surface designs. In general, the B\'ezier spline can be specified with the B\'ezier curve segments and a B\'ezier curve segment can be fitted to any number of control points. The number of control points determines the degree of the B\'ezier polynomial. This paper presents a method which determines control points for B\'ezier curves approximating segments of obtained image outline(non-parametric curve) by using the properties of cubic B\'ezier curves. Proposed method is a technique to determine the control points that has generality and reduces the error of the B\'ezier curve approximation. Main advantage of proposed method is that it has higher accuracy and compression rate than previous methods. The cubic B\'ezier spline is obtained from cubic B\'ezier curve segments. To demonstrate the various performances of the proposed algorithm, experimental results are compared.
In this paper, we defined the viseme (visual speech element) and described about the method of extracting visual feature vector. We defined the 10 visemes based on vowel by analyzing of Korean utterance and proposed the method of extracting the 20-dimensional visual feature vector, combination of static features and dynamic features. Lastly, we took an experiment in recognizing words based on 3-viseme HMM and evaluated the efficiency.