The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008 
3. RESULTS 
3.1.1 Bournemouth (node to node comparison) 
The point clouds generated from the imagery over the city of 
Bournemouth from 2005 were compared with their 
corresponding dataset from 2006 using a node to node 
comparison. Results are given in Figure 4 and Figure 5. 
Comparison of the NGATE datasets (Figure 4) highlights three 
major areas of change. Review of these areas against the image 
datasets reveals these changes to be: the redevelopment of an 
existing structure (1); the demolition of a building (2) and the 
construction of a new building (3). Other areas appear to be 
false identifications relating to errors in image correlation (at 
building edges) or vegetation growth that has been miss- 
classified at the point cloud classification stage. 
Figure 4 Results of the comparison between the 2005 and 2006 
Bournemouth dataset from the NGATE module. 
In the case of the standard Socet Set DSM (Figure 5) a greater 
level of variation between the datasets is evident. However, in 
most cases it is not conclusive whether this is due to differences 
in vegetation, surface extraction or due to actual real world 
changes. In a few cases however, it is possible to identify 
significant (large clusters of change) that do relate to real world 
changes (marked as 2, 3 and 4 in Figure 5): namely the 
demolition of a building (2), the construction of a new building 
(3) and a hot air balloon used as a viewing platform in 2005 and 
not present in 2006 (4). 
Figure 5 Results of the comparison between the 2005 and 2006 
Bournemouth dataset from the standard Socet Set DSM 
generation. 
Figure 6 gives enlarged examples of the real world changes 
found in the Bournemouth area and identified as site 1 in 
Figure 4, and site 2 found in both 
Figure 4 and Figure 5. 
Figure 6 Example real world changes in Bournemouth dataset 
(left 2005, right 2006). 
3.1.2 Heathrow (node to node comparison) 
The point clouds generated from the imagery over the area close 
to Heathrow’s Terminal Five from 2005 were compared with 
their corresponding dataset from the 2006 data using a node to 
node comparison. The results are given in Figure 7 and Figure 8. 
The significant real world changes in this test area are: the 
construction of a building (1); the demolition of a building (2); 
the erection of a new motorway gantry (3). Some of the changes 
are shown in Figure 9 and Figure 10. For both datasets the 
results do not mirror the success of the Bournemouth test site. 
In the dataset produced from the Socet Set NGATE module 
none of the three major changes are identified. In the standard 
DSM only the addition of the motorway gantry is successfully 
indentified.