Table 1. Detail of Digital Video Camera 
3.1 Robustness of Exterior Orientation in this study 
In our previous sturdy, the limitation of camera motion and 
robustness of exterior orientation in the big change of the scene 
such as the structure comer are the still remained problem. 
Figure 6 shows the example for big change of the scene. These 
video images include the near object and far object. Also, the 
yawing of camera is very quickly. 
Figure 6. Big change of the scene 
Figure 7 shows the result of exterior orientation with our old 
procedure. The trajectory of video camera is confused at the 
structure comer. The 3D coordinates of natural feature points 
are also confused. 
3.2 Application for Volume Measurement 
As the application of this study, volume measurement from 
video image sequences was performed. In this study, authors 
assume the application for volume measurement in the mining 
field. From the experience of authors, the allowable margin of 
error for volume measurement using total station in the mining 
field is about 10%. 
The exterior orientation for video camera and 3D coordinate of 
natural feature points were obtained from exterior orientation 
procedure of this study. The volume measurement from 3D 
coordinate of natural feature points was performed by using the 
modelling function of PI-3000. 
At the first, accuracy assessment for small model of gravel 
mound is described. As the next, application for real gravel 
mound is also described. 
3.2.1 Accuracy Assessment with Model Gravel Mound: In 
order to evaluate the accuracy of volume measurement, experi 
ment of volume measurement for model gravel mound was per 
formed. 
Trajectory of Video 
Camera 
Figure 9 shows the situation of this experiment. The model 
gravel mound is putted on the turntable. Therefore, video 
camera goes around the gravel mound in stable motion. 
Figure 8. Result of Exterior Orientation with Old Procedure 
Figure 9. Situation of Model Gravel Mound Experiment 
Figure 10 shows the result of exterior orientation procedure and 
Figure 11 shows the result of TIN modelling from the 3D 
coordinate of natural feature point using PI-3000.