‚works is also 
| 3D distances 
ed in different 
lual error was 
he performed 
| if scans of 
density in the 
ch unexpected 
© comparisons 
as these parts 
he Mensi data 
scanner. 
  
ignment. 
Istanbul 2004 
  
Figure 9. Errormap btw.adjacent Optech scans 
As previously mentioned, in order to generate a 3D model of 
the Chaple with adequate accuracy, we planned to choose the 
best scans among the whole data set of range data acquired 
with the four laser scanners. Unfortunately that could not be 
always done as in some cases both the Cyrax and the Optech 
sensors provided a bad intensity response, what made more 
difficult or even impossible to use the scans for the alignment. 
Moreover intensity issues of those sensors showed up in 
different ways. In the case of the Optech scanner some scans 
presented so low intensity values, where the laser beam hit the 
walls with tilted angle (i.e. no vertical scan), that no features 
could be recognizable at all (Figure 10), resulting in unuseful 
scans. 
  
Figure 10. Effect of laser beam tilting in 
Optech scans 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004 
On the other hand, the registered intensity values of Cyrax 
scans looked like to be biased, i.e. shifted towards the darker 
area of the light spectrum. In this case, for each scan the 
corresponding histogram of intensity data was computed and 
after visual analysis a constant shift has been applied, in order 
to increase the brightness of the scan at a sufficient level for 
feature matching (Figure 11 and 12). 
  
2500 
  
  
Figure 12. Modified intensity response of 
Cyrax 2500 
After scan alignment, using /mMerge module of Polyworks, 
the registered point clouds have been triangulated getting a 
unique mesh composed by several millions of triangles. At the