Comparative Study for Color Image Clustering based on Mahalanobis Distance Method

(Improved fuzzy C- mean clustering method)

Abstract

Color image has the potential to convey more information than monochrome or gray level images, RGB color model is used in many applications of image processing and image analysis such as Image Segmentation. The standard ways to deal with picture investigation and acknowledgment creatures by division of the picture into areas (articles) and figuring different properties and connections among these districts. Image segmentation algorithms, have been developed for extracting these regions. Important image segmentation is difficult to process due to the inherent noise and deterioration of the input indicators of the algorithm. The regions are not always specified, however, and it is often more fitting to consider them as fuzzy subjects in the image. In this work, an algorithm that is used to segment color images with clustering methods is defined. The lower conditions, results and results are illustrated. In the clustering algorithm, the results are compared using both Mahalanobis and Euclidean distances.