Due to some of the limitations discussed in the first sec-
tion, the proposed algorithm will do three things: (a) extract
buildings under shadows while eliminating the ambiguities
of vegetation found under the shadow area when possible,
(b) extract building candidates without the use of a DTM, and
(c) integrate the original height and spectral information for
building mask refinement, so as to prevent small buildings
from being excluded. Therefore, we accordingly proposed this
algorithmic strategy as follows: