Improving classification rates by modelling the clusters of training sets in features space using mathematical morphology operators

The exploration of features presented by the training sets of each class (size, shape and orientation) in order to construct the respective decision region borders without making explicitly any statistical hypothesis is presented in this paper. Its incorporation allows one to define more correct decision borders since there is a significant improvement in the classification rates obtained. Mathematical morphology operators are preferentially used in this methodology, which is illustrated with two spectral features (wetness tasselled cap and NDVIs vegetation index) of seven land cover classes constructed from Landsat TM satellite images of central Portugal.