Adaptive Color Image Segmentation Using Fuzzy Min-Max Clustering

Abstract

This paper proposes a novel system for color image segmentation called Adaptive color image segmentation using fuzzy min-max clustering (ACISFMC). The present work is an application of Simpson's fuzzy min-max neural network (FMMN) clustering algorithm. ACISFMC uses a multilayer perceptron (MLP) like network which perform color image segmentation using multilevel thresholding. Threshold values used for finding clusters and their labels are found automatically using FMMN clustering technique. FMMN clustering technique uses a hyperbox fuzzy set concept. In the proposed work, Fuzzy entropy is used as a tool to decide number of clusters. ACISFMC uses saturation and intensity planes of HSV (hue, saturation, intensity) color space for segmentation. Here, neural network is used to find the number of objects automatically from an image. One of the good feature of this method is that, it does not require a priori knowledge to segment a color image. The algorithm is found to be robust and relatively computationally inexpensive for large variety of color images. One application of the proposed method is demonstrated here. Experimental evaluation demonstrates the performance of ACISFMC is robust for noisy images also.