In this study, each node of the INeN represents a different class. During the training process, two
parameters are used in order to determine the number of classes: (i) The threshold value, and
(ii) the predefined value for the histogram. Low threshold values reveal the details (different classes)
in the segmented image. However, a low threshold value produces excessive number of
nodes. In order to control the number of nodes, a histogram is utilized. The histogram is obtained
by using all the pixels of the image. It gives information about the possible number of classes.