Color Image Segmentation with Different Image Segmentation Techniques
Rupali B. Nirgude, Shweta Jain · 2014
Abstract — This paper deals with different image segmentation techniques to enhance the quality of color images. The technique follows the principle of clustering and region merging algorithm. The system is combination various stages like histogram with hill climbing techniques; auto clustering includes k means clustering, the consistency test of regions, and automatic image segmentation using dynamic region merging algorithms. The different techniques of image segmentation include thresolding, clustering, region merging, region growing, color segmentation, motion segmentation and automatic image segmentation. This paper mention different methods for efficient segmentation which is combination of different algorithms. Here the given image gets converted into histogram. The histogram is graphical representation of input image. The peaks from histogram diagram are detected using hill climbing algorithm; this gives the rough number of clusters for the further steps. The clusters are form usingefficient K means clusteringalgorithm. The regions having homogenous or similar characteristics can be combining with the nearest neighbor algorithm and dynamic region merging algorithm.This segmentation technique is useful in field of image processing as well as advance medical use.