Assessment of an individual collapsed building based on the
height difference
From the matching result for the homologous points of the
building in the pre- and post-seismic stereo image pairs, the preand
post-seismic 3D ground coordinates of the corresponding
points of the building were calculated using the RPC bundle adjustment
with the best bias-compensation model. Therefore, the difference
in the 3D ground coordinates, particularly in the height,
of the corner points of the building was used to detect the status
of the building before and after the earthquake.
According to a field investigation, the height of each storey of
common residential buildings in the study area is about 3.5 m.
As a result, the number of the collapsed storeys of a building can
be estimated from the height difference of the building and storey
height. It should be noted that sometimes it is rather difficult to
identify the corresponding corner points in the pre- and postseismic
images, particularly in the post-seismic images; in this
case, the feature points on the roof of a building that are close to
the corner points in the images are measured, which might introduce
error in both calculation of the height difference and
estimation of the number of collapsed storeys.