Title :
Contiguity-constrained hierarchical clustering for image segmentation
Author :
Morales, Eduardo R Concepcion ; Mendizabal, Yosu Yurramendi
Author_Institution :
Fac. of Inf., Univ. of Cienfuegos, Cienfuegos, Cuba
Abstract :
Traditional clustering methods do not take into account any relations possibly present in data. This paper introduces a contiguity-constrained algorithm with an aggregation index which uses neighbouring relations present in the data. Experiments show the behaviour of the proposed method in the case of medical image segmentation.
Keywords :
biomedical MRI; image segmentation; medical image processing; pattern clustering; aggregation index; contiguity-constrained hierarchical clustering; medical image segmentation; neighbouring relations; Clustering algorithms; Image segmentation; Indexes; Magnetic resonance imaging; Manuals; Partitioning algorithms; Pixel; Constrained hierarchical clustering; Image processing applications; Image segmentation;
Conference_Titel :
Image Processing Theory Tools and Applications (IPTA), 2010 2nd International Conference on
Conference_Location :
Paris
Print_ISBN :
978-1-4244-7247-5
DOI :
10.1109/IPTA.2010.5586724
