MATLAB 2010B and its Neural Network, Image
Processing and Statistics Toolboxes are used for performing
all computations and simulations presented in this paper.
Three major experiments are conducted. In the first one,
classifiers are trained using all the extracted features without
any statistical pre-processing. In the second, the extracted
features are normalized before being fed to a classifier and in
the last one, the trained classifiers use features obtained after
pre-processing with PCA.