The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008 
2.4 Object Space Control Frame 
The 3D object space control is an important part in the 
development of craniofacial spatial data acquisition system. The 
3D object space control consists of a special designed chair with 
adjustable head rest and high accuracy photogrammetric control 
frame. The photogrammetric control frame used to provide 
high accuracy 3D ground control to the stereo images via 
signalized retro-reflective targets. Figure 12 shows the side 
view of the photogrammetric control frame. 
Figure 12. Object space control frame 
The photogrammetric control frame requires photogrammetric 
calibration in order to determine the 3D coordinates of the paper 
targets. To calibrate the control frame four or more convergent 
photographs are taken with a high precision invar scale bar 
placed in the middle of the control frame. A bundle 
adjustment process is needed to determine the 3D coordinates of 
the targets. It is not necessary to have any previous known 
control point in the adjustment as in the case of an absolute 
orientation of a stereo-model. During data collection process, 
the patient will sit down on the chair with the head placed at the 
middle of the photogrammetric control frame. The level of the 
eye is parallel to x-axis and z-axis (Figure 13). 
The photogrammetric control frame consists of 39 control 
points with 6mm diameter paper targets and 6 photogrammetric 
coded targets arranged in grid form. The big number of 
control points is fully needed to increase the accuracy of relative 
orientation of the stereo images, where all the points were 
digitized accurately. The control frame can be moved 
precisely using built-in moveable gear along the y-axis. 
CJ Cl 
Figure 14. Image datasets from camera system 
Figure 13. The position of patient’s head during data acquisition 
2.5 The Craniofacial Raw Datasets 
As mentioned elsewhere in the report, the data collection task 
involved the collection of two types of craniofacial spatial data, 
the stereo images and the 3D laser scan surfaces. This mean 
that for each individual that involved as samples or/and 
populations was photographed using stereo camera and scanned 
by using the laser scanners. The data collection process was 
done one after another where most of the cases, the stereo 
images were captured first because the data collection period of 
the stereo camera system was very fast (0.2 mili-seconds) 
compare to the scanning period which took about 19 seconds to 
complete the scanning process from two scanners. A complete 
datasets will consist of three stereo images, two convergent 
images and two 3D surfaces (which also known as shells). 
Figure 14 shows the image datasets, while Figure 15 shows the 
laser scanner dataset. 
As seen in both Figure 14 and Figure 15, the photogrammetric 
control frame image was included in the raw datasets. 
Figure 15. 3D laser scanner raw datasets - (a) right shell, (b) left 
shell