Color image segmentation is still a challenging problem. Literature reveals many supervised algorithms wherein the primary input is the number of segments to which the image is to be segmented. Currently researchers are focusing on un-supervised segmentation algorithms. The main advantage of the proposed method is that no a priori information is required to segment the given color image and hence considered as an unsupervised approach. The proposed method is found to be reliable and works satisfactorily on different kinds of color images. Subjective comparison and objective evaluation shows the efficacy of the proposed method over other existing methods. © 2010 ACM.