Applying Functional Networks to Fit Data Points from B-Spline Surfaces

Recently, a powerful extension of neural networks. the so-called functionaI networks. has been introduced / I /. This kind of networks exhibits more versatility than neural networks so they can be successfully applied to several problems in Cotputer-Aided Geometric Design (CAGD). As an illustration, the simplest functional network representing tensor product surfaces is obtained. Then, functional networks' formalism is advantageously used to fit given sets of data from B-spline surfaces through a Btzier surface. The proposed method also determines the degree and the coeflcients (the control points) of the approxitnating surface thatfits the given data better. This new approach is very general and can also be applied to any other interesting family of approximating basis functions in CAGD.