On a Granular Approach for Fuzzy Color Modelling.

In this paper, a new fuzzy granular approach for modeling color categories is proposed. Fuzzy granular colors where introduced by some of the authors as a way to model color categories with non-convex membership functions, particularly those categories that can be defined in terms of the union of a finite collection of sub-categories (which are very commonly found and applied in practice). To build a granular color, an aggregation of fuzzy colors having semantic relationships, the so-called \"granules\", is performed. In this paper we introduce the use of Voronoi tessellations for modelling the granules, comparing its behavior with respect to a sphere-based approach. We illustrate the advantages of our approach with respect to current state of the art on the basis of some experiments.