A New B-spline Curves Fitting Algorithm for Scatter Data

In this paper, a new curve-fitting algorithm is presented. This algorithm can automatically fit a set of scatter data points with optimal degree of piecewise B -spline curves. In the first step,knot points are identified from the given data set and are further classified as either corner points or local maximum curvature points. Then,we obtained the optimal degree of piecewise B-spline curves by Bayesian decision rule. Curves fitting is in the last step. This algorithm has been tested and can be applied to inverse engineering and image curves fitting.