COLLAGE-BASED INTERPOLATION OF REAL-DATA SETS

Abstract

This paper presents a recurrent fractal interpolation method (approach) for one-dimensional sets of real-data. The method explores both the local collage idea, developed originally for image compression purposes, and the basic platform for generating of non-recurrent fractal interpolation functions – attractors of iterated function systems (IFS). The characteristic feature of the developed approach – the recurrent fractal interpolation functions are obtained by applying specialized correction procedures to the approximants of the real-data sets, i.e. to the attractors of local IFS, generated using self-similarities detected within the data under processing.