The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008 
3. CRANIOFACIAL SPATIAL DATA REDUCTION 
SYSTEM 
The data reduction system involved with the method used to 
process the raw data (as acquired using data acquisition system). 
The method used consisted of pre and post processing of the 
raw data. There are two types of data pre and post processing 
tasks involved in the project. The first pre and post processing 
task involved the processing of image datasets (as acquired 
using camera system), while the second one involved with the 
processing of 3D laser scanner point clouds data. The details 
of the data processing are as below: 
3.1 Pre and Post Processing of Images Acquired from the 
Camera System 
The acquired stereo and convergent images involved with two 
pre-processing tasks. The first task involved with 
photogrammetric triangulation process. The aim of the process 
is to measure the 3D coordinates of the natural landmarks on the 
craniofacial surface. The 3D coordinates were than use as 
control points in photogrammetric absolute orientation process 
(Figure 16 and Figure 18). 
Figure 16. The flowchart to show the measurement of natural 
features 3D XYZ coordinates using Photogrammetric 
Triangulation process 
Figure 17. The flowchart to show the stereophotogrammetric 
measurement process of stereo images using DVP 
Photogrammetric System 
Figure 18. Measurement of natural features using convergent 
photogrammetric method 
The stereo images were than proceed with the photogrammetric 
stereo orientation which involved interior, relative and absolute 
orientation process to generate the stereo model. The stereo 
vectorization was than applied to digitize the 3D XYZ 
coordinates of the craniofacial landmarks (see Figure 17 and 
Figure 19). The accuracy of the stereo digitizing process was 
evaluated using the RMS value of the absolute orientation. Most 
of the cases in craniofacial mapping required the RMS value of 
0.5mm for X, Y and Z coordinates, respectively. 
Figure 19. Post processing of stereo images using DVP system. 
3.2 Pre and Post Processing of 3D Laser Scanning Datasets 
The data processing of the 3D laser scanning datasets involved 
with six common 3D laser modeling process which consists of 
filtering noise, initial registration and fine registration of the 
two shells, merging, holes filling and smoothing (as in Figure 
20 and Figure 21). The common processing steps mentioned 
above was offered by most of the laser scanning data processing 
software such as RapidForm 2004 (INUS Technology, Korea) 
and Polygon Editing Tools (PET) software (Konica Minolta, 
Japan). 
Figure 20. The flowchart show the pre-processing steps to 
process the 3D laser scanning datasets 
The post processing involved with the measurement of 
craniofacial landmarks on the 3D craniofacial surface model. 
The process required the user to identify the location of the 
landmarks on the 3D surface based on the terrain (DTM) of the 
surface and the break-surface, and finally the location was 
digitized. RapidForm 2004 software offered auto-measure 
function to measure slope distance, along surface distance and 
angle between the selected landmarks.