1181
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008
Figure 7 Results of the comparison between the 2005 and 2006
Heathrow dataset from the NGATE module.
Figure 8 Results of the comparison between the 2005 and 2006
Heathrow dataset from the standard Socet Set DSM generation.
Figure 9 Example real world changes in Heathrow dataset (left
2005, right 2006).
Figure 10 Example real world changes in Heathrow dataset (left
2005, right 2006).
3.2. Bournemouth/Heathrow (point to node comparison)
A second set of test results were run using the point to node
routine described previously, comparing 2005 NGATE point
cloud data against an octree derived from 2006 point data, and
vice versa. The closest point was located and the elevation
difference between the two points was computed. A threshold
for the elevation difference was defined manually, consulting
the data and determining the level at which an actual difference
in elevation value indicated a possible real world change. Given
the nature of the change being considered
(demolition/construction of new buildings at least 5m in height)
this was set at a threshold of 4.5m. Elevation differences greater
than this value are displayed in Figure 11. Note: whereas in the
node to node comparison octrees were generated with a
maximum point per node value of 150 points, the octrees used
in the point to node comparison limited the point per node value
at 50, thereby deriving octrees of a higher resolution.
Figure 11 A point to node comparison highlighting change
between 2005 and 2006 Bournemouth datasets.