Defining the fuzzy transform on radial basis

Abstract

Fuzzy transform (F-transform) is a functional operator that proved to be effective in multiple processing tasks concerning time series and images, including summarization, compression, filtering and information fusion. Its operational definition is originally based on Ruspini partitions that, if on one side provides an easier interpretation of results, on the other it might limit the application of F-transform for higher dimensional data spaces. In this paper we propose a radial definition of F-transform that is based on the distance between points in a p-normed space.