Color Based Image Segmentation Using Fuzzy Logic

This paper projects to perform Image segmentation. Segmentation of an image can be done in several ways. Here, pixel based or color based technique is used to fragment an image. The rule based fuzzy logic system is applied to the image to segment its parts based on the values of each of the pixel (RGB values). In this paper, firstly the features of fuzzy logic reasoning are explained. Then, a fuzzy inference system is designed by defining various fuzzy rules and fuzzy sets as necessary. Afterwards, an algorithm is developed to scan the image for its various color components categorizing its colors fundamentally. The idea is to input the image, scan it for various color components, and then provide this data to the FIS system to obtain a constant value output. This divides the image in color lots and hence segmenting it. The segmented images are then shown, along with their computational time.