Segmentation of Colour Images by Modified Mountain Clustering
Keywords: Fuzzy C-means clusters, computer vision, fuzzy Gaussian function, fuzzy image p&cessing, membership function, validity criteria, mountain clustering
AbstractSegmentation of colour images is an important issue in various machine vision and image processing applications. Though clustering techniques have been in vogue for many years, these have not been very effective because of problems like selection of the number of clusters. This problem has been tackled by having a validity measure coupled with the new clustering technique. This method treats each point in the dataset, which is the map of all possible colour combinations in the given image, as a potential cluster centre and estimates its potential wrt other data elements. First, the point with the maximum value of potential is considered to be a cluster centre and then its effect is removed from other points of the dataset. This procedure is repeated to determine different cluster centres. At the same time, the compactness and the minimum separation is computed amongst all the cluster centres, and also the validity function as the ratio of these quantities. The validity function can be used to choose the number of clusters. This technique has been compared to the fuzzy C-means technique and the results have been shown for a sample colour image.
How to Cite
Jha, D., & Hanmandlu, M. (2002). Segmentation of Colour Images by Modified Mountain Clustering. Defence Science Journal, 52(3), 293-302. https://doi.org/10.14429/dsj.52.2184
Special Issue Papers
Copyright (c) 2016 Defence Science Journal
Where otherwise noted, the Articles on this site are licensed under Creative Commons License: CC Attribution-Noncommercial-No Derivative Works 2.5 India