Color image analysis using fuzzy set theory

An application of fuzzy set theory to the color image analysis is presented in this paper. It is developed to inference the weather state from a color image. For a color image, its RGB components are first converted into HSI components. For each component, a 3-layer quadtree is created for image analysis. Then the histogram of each block image from the top layer to bottom layer is computed. The histogram is further converted into a membership function by using the scheme of a 6-parameter fuzzy number. The fuzzy inference for the given image is thus achieved by using the generated membership functions.