The effect of different consequent setting on the effectiveness of Kóczy-Hirota fuzzy rule interpolation in colonoscopy image classification

The effect of the refinement of the consequent categories in K�czy-Hirota fuzzy rule interpolation is studied in the model system of colonoscopy image classification. The colonoscopy image segments are classified based on 99 antecedents, which are statistical parameters of the image segments. The number of consequents is increased from 2 - contains polyp, or does not contain polyp - to a more refined set distribution based on the polyp content of the image segment.