Home Image Processing Cluster Analysis Fuzzy cMeans Clustering  
See also: Hierarchical Cluster Analysis, kMeans Clustering


Fuzzy cMeans Clustering 

Fuzzy cmeans clustering has been introduced by Bezdek et al. in 1984. It is an unsupervised algorithm which requires a userdefined number of clusters. The difference to classical kmeans clustering is based on the idea of assigning a class membership value to each pixel which reflects the probability of belonging to a particular class. Thus fuzzy cmeans delivers class regions which overlap in the image. The cluster analysis is performed using the loaded spectral descriptors, which may be (de)selected on an individual basis by ticking off the corresponding check box in the list of spectral descriptors. The results of the kmeans clustering is displayed as a colored map, which can be processed in the usual way (copy to the 2D Imager, copy to the image stack, export to DataLab). Further, a particular class of pixels can be copied into the mask editor, thus enabling the user to exclude these pixels from further calculations. The fuzzy cmeans algorithm is prone to unfavorable initial positions of the cluster prototypes. This can be circumvented by repeating the algorithm with different starting positions several times and finally using the set of starting conditions which result in the smallest intracluster distance. If you tick off the check box "Optimize", the fuzzy cmeans model is calculated 20 times, showing the best of these trials.
The optimal setting of the fuzzy weight parameter is not known a priori and must be found empirically. A simple tool for this is checking the pixel purities while scanning the weight parameter (see tabsheet "Scan Fuzzy Weight").
