6B-4-3 
At first step, recorded image was converted to some of stereo 
pair digital imageries by Video/Computer interface. As 
Video/Computer interface, both image capture board (MTATCL) 
and Digital Image Recorder (MR-1500) were used in order to 
verify its recorder's performance on 3D-measurement accuracy. 
At second step, each center position of circle (e.g. Center of 
gravity) was estimated as its area gravity for each of imagery by 
basic image processing procedure that illustrated on Fig. 5. 
Original Image Binary Image Area Gravity 
Fig. 5 Processing Procedure for Area Gravity Estimation 
And each of its position of gravity on image coordinate was 
calculated following equations: 
where, (Uc,Vc): the Position of Area Gravity, 
(Ui,Vi): Image Coordinate, Wi:Weight 
At final step, the orientation of single photograph was carried 
out each left and right imagery by the bundle adjustment using 
mentioned 9 point's positions. And 3D-measurement was also 
performed. 
2) Results of Experiment 
The theoretical accuracy for verification is calculated 
following equations: 
a xy = (H/f)o p (3) 
Gz = (H/f)(H/B)ff p (4) 
whwre, FI; Distance from Camera to Target(Altitude), 
B; Base-Line, f; Focal Length, 
a p ; Pointing Accuracy 
Table 1 is the results of this 3D-measurement experiment. 
Table 1 Comparison of Theoretical Accuracy 
B 
(mm) 
Digitizing 
Method 
Theoretical Accuracy 
a xy(mm) 
a z(mm) 
300 
Capture Board 
0.100 
0.700 
300 
Digital Recorder 
0.100 
0.900 
100 
Capture Board 
0.100 
2.000 
100 
Digital Recorder 
0.100 
2.700 
In the use of capture board, sufficient results were obtained. 
On the other hand, Digital Recorder could not provide enough 
results. One reason that we suppose is some of estimated 
parameters such as scale factor were used in the inner 
orientation. 
The reason why no difference of accuracy was identified 
depend on digitizing, is seems to be the method how center 
position was estimated. 
In this experiment, in order to identify the capability of sensor 
integration, 3D sensor data was used as camera angles and also 
executed 3D-measurement. Its results had almost same accuracy 
as in the use of parameters that introduced by the internal 
orientation. 
Although some unsolved problems have still remained, our 
development was verified to be suitable. 
3. UNDERGOING AND FUTURE RESEARCH 
When authors had started this research, we had already two 
kinds of application plans at the same time. Next two topics are 
the related research projects based on this development. 
3.1 Fixed Point Observation as Total Monitoring System for 
Forest Ecosystem 
Fig.6 shows our system design of canopy crane in Kuba 
National Park, Sarawak, Malaysia. In the end of this year, it will 
be constructed. In this research project, the meteorological survey 
and many kinds of ecological surveys should be done. This 
system aims to realize virtual field. When this plan will be 
realized, every research scientists will be able to monitor actual 
field phenomena from any places through Internet. 
Video Monitoring System 
for Forest Ecosystem 
3.1 Moving Observation as Mobil Mapping System 
The authors have been repeating the experiment on mobile 
mapping system. Fig.7 shows an actual experiment picture at 
riverside road. This system aims to draw the map as the result of 
spatial data acquisition. 
Through previous experiments, every observation instruments 
was identified to be sensitive for car peculiar oscillation. 3D 
sensor was most sensitive especially for an azimuth direction. 
Then we have been revising the system and considering the 
method how we have to measure. And at the same time, 
preventive mount for its damage has been designing and 
developing. These mentioned preliminary experiments have 
almost completed. In near future, we should start substantial 
experiments using developed mount for putting all instruments 
on.