A Robust Spline Filter on the basis of L2-norm

Abstract We aimed to develop a sort of Robust Spline Filter that is much less sensitive to outliers than the ordinary Spline Filter. This new Robust Spline Filter is based on the concept and definition of the Spline Filter that was introduced by M. Krystek. This newly developed Robust Spline Filter inherits all the advantages directly from the ordinary Spline Filter. The L2-norm based robust algorithm is proposed and the computational procedure is described and illustrated. By applying this Robust Spline Filter to several measured datasets, its effectiveness was demonstrated in contrast with the ordinary Spline Filter. This new filter can be also developed for any continuous plane curve and space curve based on the one-dimensional curve theory.