Title :
Exponential-Distance Weighted K-Means Algorithm with Spatial Constraints for Color Image Segmentation
Author :
Hung, Wen-Liang ; Yang, Miin-Shen ; Hwang, Chao-Ming
Author_Institution :
Grad. Inst. of Comput. Sci., Nat. Hsinchu Univ. of Educ., Hsinchu, Taiwan
Abstract :
The weighted k-means proposed by Huang et al. (2005) could automatically calculate feature weights. On theother hand, the fuzzy c-means (FCM) with spatial constraints(FCM_S) is an effective algorithm suitable for imagesegmentation. In this paper we propose a robust exponential distance weighted k-means (EDWkM) algorithm with spatialconstraints. The proposed algorithm is applied in color imagesegmentation. Several numerical and color image experimentsare performed to compare the EDWkM with existing methods.Experimental results show the effectiveness and superiority of the EDWkM algorithm.
Keywords :
fuzzy set theory; image colour analysis; image segmentation; pattern clustering; color image segmentation; exponential distance weighted k-means algorithm; fuzzy c-means with spatial constraints; spatial constraints; Clustering algorithms; Color; Image segmentation; Noise measurement; Partitioning algorithms; Pixel; Robustness; Clustering; Color image segmentation; Exponential distance; K-means; Weighted k-means;
Conference_Titel :
Multimedia and Signal Processing (CMSP), 2011 International Conference on
Conference_Location :
Guilin, Guangxi
Print_ISBN :
978-1-61284-314-8
Electronic_ISBN :
978-1-61284-314-8
DOI :
10.1109/CMSP.2011.33
