Granular Modeling of Fuzzy Color Categories

In this paper, we introduce fuzzy granular colors for modeling color categories. Fuzzy granular colors are built by aggregating fuzzy colors having semantic relationships with regard to a certain color category. Our proposal allows us to model color categories which comprise disjoint fuzzy subsets of colors, as well as those having a nonconvex representation in the color space, among other advantages. Such categories are used very often by humans in different real contexts. Fuzzy granular colors are appropriate to provide color models able to deal with ill-defined boundaries, subjectivity, and context-dependence. We illustrate the advantages of our approach with respect to current state of the art with several experiments.