Fuzzy texture detection using homogeneity cues

A fuzzy inference approach for visual texture detection is discussed. Pixel values in a homogeneity image are fed into a fuzzy inference engine in order to classify each pixel into regions labeled as "homogeneous" or "texture". An in-depth analysis where different grades of texture homogeneity are also feasible was presented. The performance of the proposed method illustrated on some sample images.