Color comparison in fuzzy color spaces

Abstract

In this paper we study the problem of color comparison in computers in the setting of fuzzy color spaces. First, we study resemblance relations between precise colors induced by fuzzy colors and fuzzy color spaces. Such resemblance relations are equivalent to fuzzy categorizations of color spaces, that are crucial in order to capture the human's perception of color. Second, we consider the case of color information expressed by means of fuzzy colors used either in a conjunctive or disjunctive way. In order to match pieces of color information we use concepts of resemblance, inclusion, compatibility, and possibility/necessity between the fuzzy colors involved in the definition of different pieces of color information, using well known results from the fuzzy set and possibility theories. Finally, we define inclusion/similarity indexes for comparison of fuzzy colors, including a novel approach to calculate inclusion and overlapping between fuzzy colors on the basis of quantification techniques.