Title :
A novel Fuzzy C Means algorithm based on distance modification for image segmentation
Author :
Naixiang Li ; Peng Guo
Author_Institution :
Dept. of Comput. Sci. & Inf. Eng., Tianjin Agric. Univ., Tianjin, China
Abstract :
A novel Fuzzy c Means(FCM) algorithm with modified distance computation is proposed in this paper. We modify the distance in FCM with the neighborhood information of cluster centers. The distance in FCM is composed of the Euclidean distance and a characteristic distance, and the characteristic distance is calculated with a pixel center window and tuned with a coefficient. The Gamma function is selected to generate coefficients in our works. Experimental results show high performance of our approach.
Keywords :
fuzzy set theory; image segmentation; pattern clustering; Euclidean distance; FCM; Gamma function; cluster centers; fuzzy c means algorithm; image segmentation; modified distance computation; neighborhood information; pixel center window; Clustering algorithms; Euclidean distance; Image segmentation; Indexes; Linear programming; Magnetic resonance imaging; Signal processing algorithms; Characteristic Distance; Distance Modification; Fuzzy C Means; Image Segmentation;
Conference_Titel :
Fuzzy Systems and Knowledge Discovery (FSKD), 2013 10th International Conference on
Conference_Location :
Shenyang
DOI :
10.1109/FSKD.2013.6816188
