The work here presented concerns an application study on the
process of urban sprawl, carried out by the use of remote sensed information,
from ASTER data which already proved to be suitable for these
topics. In order to produce synthetic maps of urban areas of the territory,
ASTER images were classified using two automatic classifiers, which
areMaximumLikelihood Classifier (MLC) and Support VectorMachines
(SVMs), the latter based on machine learning theory (Zhu, and
Blumberg, 2002). The aimwas to compare performances in terms of robustness,
speed and accuracy of the two classifiers, regarding urban
pixels.