ImLab Scilab Function
Last update :

imhierarclustrain - hierarchical clustering method, training phase: second step of image segmentation using hierarchical clustering method.

Calling Sequence

[prototype] = imhierarclustrain(learningdata,classnb)

Parameters

Description

Thanks to learning data and following hierarchical clustering method, this function creates classes and returns their features in a prototype. The number K of classes is user defined.

Algorithm :

1. Initialisation: each cluster contains a single object.

2. Merge the two closest clusters. The new cluster is represented by its centre of gravity. Step 2 is repeated until the number of clusters equals K.

The result prototype is formed with the K centres. Note that the distance used is the euclidean one.