Measuring Ambiguities In Images Using Rough And Fuzzy Set Theory

Images, in general are ambiguous in nature. In this paper, we propose the combined use of rough and fuzzy set used to capture the indiscernibility among nearby gray values, whereas fuzzy set theory is used to capture the vagueness in the boundaries of the various regions. A measure called rough-fuzzy entropy of sets is proposed to quantify image ambiguity using which a characteristic measure of an image called the average image ambiguity (AIA) is presented. The rough-fuzzy entropy measure is used to perform various image processing tasks such as object / background separation, multiple region segmentation and edge extraction, and the corresponding performance are compared to those obtained using certain existing fuzzy and rough set theory based image ambiguity measures. Extensive experimental results are given to demonstrate the utility of measuring image ambiguity using the proposed rough-fuzzy entropy measure.