Change detection results indicate that different polarimetric
parameters can provide complimentary information about damaged
areas. Consequently, the performance of change detection can be
further refined by merging multiple polarimetric parameters. In
addition, spatial contextual information is also important in further
refining the performance of change detection. The Markov Random
Field (MRF) model can provide a useful tool for characterizing contextual
information and a methodological framework that allows
the different information to be merged in a consistent way (Geman
& Geman, 1984).