INTERMEDIATE FUZZY NUMBERS FOR COLOR KNOWLEDGE REPRESENTATION

Abstract

Color knowledge representation is an essential aspect for designing intelligent graphic systems. Since the mapping between low-level color numerical values and high-level semantics is crucial, the theory of fuzzy numbers is greatly needed for representing color concepts. The limited linguistic color words bring about the requirement for representing intermediate color concepts. However, there is not corresponding definitions in the fuzzy realm. To overcome this representational inadequacy, this work addresses the issue of formally defining the intermediate fuzzy concept. Since definitions of the intermediate fuzzy numbers need a well-defined order between two basic fuzzy numbers, this paper first presents a cognition based linguistic ranking method. Then two novel definitions, namely, intermediate fuzzy numbers and intermediate operations on fuzzy numbers, are presented. The definitions are of abstract ones in order for general use. Besides, the paper also presents and studies some specific intermediate fuzzy numbers and intermediate operations, which are important for modelling gradual changes between two color terms in Color Naming System.