Fuzzy Rule-Based Skin Color Modeling and Classification

Abstract

Researchers have pursued skin color modeling as a means for classifying skin color pixels mainly throughhistogram method and Gaussian Model. This paper reports the modeling of a skin color classifier using Fuzzy Logic. The merit of modeling the skin color via Fuzzy Logic is the easily available platform offers by Fuzzy Logic to adapt the classifier. The fuzzy rules for the classifier are obtained through a tree-like selective procedure from the 3D-histogram of the RGB color space. Classification results pertaining to the method is presented. Some concluding remarks are derived towards the end of the paper.