6B-5-6 
(1) experimentation 
(a) Establishment of subjects 
Fig.5-2 shows the subjects of measurement points and 
control points. These subjects measure 20cm x 20cm,and 
are made of plastic board. The three-dimensional positions 
of subjects were located by the method of intersection and 
resection by utilizing a theodolite. 
(b) Photographing 
Experimentation was carried out for three cases as fol 
lows. 
Case T.Photographing distance is 5m,Base-height ratio is 
1/2.5 
Case 2:Photographing distance is 4m,Base-height ratio is 
1/2 
Case 3:Photographing distance is 4m,Base-height ratio is 
1/4 
z 
Fig.5-2:Subjects of measurement points and control points 
(2) Result of analysis and consideration 
For the accuracy test, the authors evaluated the three- 
dimensional measurement by utilizing photograph and 
theodolite. Process of analysis is shown as follows: The 
first defined inner orientation parameters of camera using 
control points in Fig.5-2. The secondly orientated camera 
under the information on inner orientation parameters and 
exterior orientation parameters that were obtained by 
using an inertial device. The taken photos were enlarged 
about 20cm x 20cm,and the photographic coordinates 
were measured by a digitizer whose resolution is 0.025mm. 
The number of selected control points are 18 points. The 
measurement points and the control points are not the 
same, but the points from case 1 to case 3 are the same 
points. Result of analysis are shown in Table 5-1.From 
these results, the average of standard deviation of meas 
urement error on inertial photogrammetry system is 
0.019m for x, y plane and 0.073m for z depth. The cause 
of error will still give incomplete measurement accuracy of 
exterior orientation parameter by inertial device, and dis 
agree with center of camera lens and inertial device, etc. 
Table. 5-1: Results of analysis 
Case 1 
Actual measurement 
Normal" 
IPS" 
Photographin g 
Left photograph 
5.019 
4.953 
5.019 
distance H(m) 
Right photograph 
5.019 
5.035 
4.966 
Base length B(m) 
2.000 
2.039 
1.94 7 
Base-height ratio B/H 
1/2.5 
Measurement error’ 3 
X(m) 
0.002 
0.023 
Y(m) 
0.001 
0.012 
Z(m ) 
0.009 
0.114 
Standard deviation of 
measurement error 
X(m) 
0.004 
0.003 
0.025 
Y(m) 
0.004 
0.010 
0.016 
Z(m) 
0.011 
0.013 
0.112 
AZ/Z 
1/550 
1/44 
Case2 
Case3 
Actual measurement 
Normal" 1 
IPS'* 
Actual measurement 
Normal* 1 
IPS" 
4.019 
3.971 
4.082 
4.019 
4.016 
4.070 
4.019 
4.003 
4.022 
4.019 
3.995 
4.041 
2.000 
2.021 
1.917 
1.000 
1.026 
1.092 
1/2 
1/4 
0.002 
0.025 
0.002 
0.037 
0.001 
0.017 
0.001 
0.014 
0.008 
0.083 
0.009 
0.072 
0.004 
0.002 
0.019 
0.004 
0.003 
0.019 
0.004 
0.001 
0.021 
0.004 
0.001 
0.01 6 
0.007 
0.010 
0.079 
0.014 
0.013 
0.029 
1/500 
1/50 
1/440 
1/55 
*1 Normal:Normal photogrammetric method. 
*2 IPS: Inertial photogrammetry system. 
+ 3 Measurement error:Absolute value of measured value by photogrammetry 
from measured value by theodolite. 
Calculate equation of actual measurement cr XY = (H/f) • Op 
o 2 -’(1 /0) * o p ( = 0. 028m C7 P =0. 025mm 
Note: Me asurements error of Normal.lPS and AZ/Z are average of measurements 
of 10 measurement points. 
6.Conclusion 
(1) This paper’s method has some advantages for method 
using GPS; for example, this method makes possible local 
positioning system and photogrammetry without using any 
control points in indoor, underground and among the 
mountains that cannot catch a radio wave from GPS satel 
lite. 
(2) The device and measurement accuracy have to be 
further improved; however, as the results of the experi 
ment, a lot of basic data for developing the research could 
be collected. 
(3) When we do the positioning using accelerometer and 
gyroscope , we think that relative accuracy of the result 
obtained by movement (The distance is longer and the 
difference of elevation is larger.) with some amount de 
gree’s of movement improves. 
(4) lt is necessary to examine the method not to join the 
outside power like the vibration and the impact, etc. to the 
inertial device. 
(5) Measuring time should be shortened as much as possi 
ble. Moreover, it is thought examining the measurement 
method is necessary so that ending the measurement on 
the way ( If possible, an already-known point is desirable.) 
and starting the measurement from the place again can be 
done. 
(6) This paper showed that it is possible to get exterior 
orientation parameter, that is to say information on the 
position and attitude of camera, by using an accelerometer 
and vibration gyroscope. 
(7) This paper developed photogrammetry without using 
any control points.