Analysis of Color Image Segmentation by K-means Clustering

Abstract: Image segmentation is a crucial step in image processing. The purpose of image segmentation is to divide the image into objects and extract the useful information. The level of segmentation is application dependent. Color image enhances the process of feature extraction and matching as compared to grey level image. There are various techniques of segmentation of color images but the method of clustering by K-means algorithm is discussed in this paper. K-means clustering is the simplest method of clustering the objects. The number of clusters to be partitioned and a distance metric to quantify how close two objects should be to each other must be specified in the algorithm itself. The paper shows the various results of k-means clustering based on objects as well as colors by using the MATLAB software.

Keywords: Segmentation, K-means, image processing, Clustering, Pixels, Information, MATLAB.