Technological advances and the growing availability of highspatial
resolution
imagery
has
focused
attention
on
the
limitations
of
traditional
pixel-based classification
methodologies
and
precipitated
the need for
more advanced
methods. Object-based
classification
methodologies,
that
take advantage
of
both the
spectral
and contextual
information
in
remotely
sensed
imagery,
may
hold
the
answer.