Color Image Segmentation by EFCM Clustering (using Mahalanobis distance)


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 approaches to image analysis and recognition beings by segmentation of the image into regions (objects) and computing various properties and relationships among these regions. Image segmentation algorithms, have been developed for extracting these regions. Due to the inherent noise an degradation of the input cues to the algorithm , meaningful image segmentation is difficult process. However, the regions are not always defined, it is sometimes more appropriate to regard them as fuzzy subjects of the image. In this work the way is described an algorithm, which are used to segmentation of color images with clustering methods. This algorithm is tested on ten different color images, which are firstly transformed to R*B*G* color space. Conditions, results and conclusions are described lower. The results are compared using both Mahalanobis and Euclidean distances in the clustering algorithm.


Image, Segmentation, EFCM