Advanced Image Segmentation Technique using Improved K Means Clustering Algorithm with Pixel Potential

Abstract

Image segmentation is the method of partitioning, or segmenting, different parts of the image in such a way that all segments are disjoint and each has similar elements. This process has wide applications in the field of medicine and photography industry. There are many ways in which image segmentation can be performed, from which K-Means clustering algorithm is well renowned due to its simplicity and effectiveness to perform the task. In this paper, an improved variant of K-Means Clustering algorithm is presented. The algorithm rests on applying partial contrast stretching, eliminating randomness in choosing the initial cluster centres for K-means algorithm, and removing the unwanted noise from median filters to obtain a high-quality image output.