l al
i a i 2 a i 3’
/cost))
0
-sin<j> >
/ COSK
sinK
0 1
l 1
0
0 \
a 2
ia 2 2a 2 3
r
0
1
0
-SinK
COSK
0
0
COSO)
sinoo
\a 3
i &3 2 a 3 3J
isint))
0
COSij) /
\ 0
0
1/
\0
-sinto
COS 00/
a 11= cos<i>cosK
ai2= coscfisinKCOsa)
ai 3 = coscj)sinKsinto-sincj)cosu)
a 2 i=-sinK
a 22 = coskcosw
a 2 3= cosKsinu
a 3 1= sin<J)COSK
a 32 = sinijjsinKcosuj-cos^sinco
a 33 = sin4>sinKsina)+cos(j)coscj
Eighteen unknown parameters of Xoo, Yoo, Zoo, Xi, Yi, Zi ,oi 0
<f>o,K 0 ,wi / c}>i,K 1 ,a) 2 ,cj) 2 ,K 2/ a) 3 ,<f> 3 ,K 3 are determined by the known
control points, and/or measured attitude and tracking data.
Following three methods of exterior orientation were compared.
a. Independent Solution
Ground control points are applied to either fore or aft
scene.
b. Simultaneous Solution
Ground control points are applied to both fore and aft
scenes.
c. Integrated Solution
Ground control points as well as satellite data of
attitude and tracking data with the respective weights
are applied to both fore and aft scenes. The advantage
of this solution is to be able to reduce the number of
ground control points. Accuracies of satellite data
between a.Landsat SIAT data and b.G.P.S. data were
analysed. The result in assumption of LANDSAT SIAT
data was shown.
Table 1 shows the comparison between the given simulated
parameters and the determined parameters for the test area.
As the results of the above test, accuracy to determine the
exterior orientation parameters is much better in the case
of simultaneous solution rather than in the case of inde
pendent solution. Table 2 shows the mean errors of position
and height for 79 sampled points of which 14 points were
used as control points.
AUTOMATED THREE DIMENSIONAL MEASUREMENT
Automated three dimensional measurement is of importance for
automated stereomapping. Following three types of auto
correlation to recognize the corresponding conjugate point
were compared with respect to mismatch ratio and computing
time.
a. Fourier adjustment
b. Maximum correlation coefficient
c. Minimum absolute sum
487
