Subtractive clustering: A tool for reconstructing noisy curves

A new approach for reconstructing noisy curves has been proposed in this paper. Direct use of curve fitting on the noisy data yields very poor results and so some form of smoothing is required to eliminate this noise. Subtractive clustering combined with traditional curve fitting has been used to generate the original curves. The curves reconstructed here are the Bezier curves which are a class of curves used for ab initio designs.