After the feature extraction process, vectors are presented to the artificial neural networks for the
training. In the training of the INeN, the number of nodes of the network is automatically determined
by its unsupervised learning scheme. A label is assigned to each node and labels of all nodes are
saved in an index layer. During the segmentation process, a feature vector is formed for each pixel
and presented to the INeN. The pixel under consideration is labeled by using the label of the network
node, which is the nearest to the feature vector.