rom
red
tal
xel
the
the
ial
was
the
age
sor
ess
of
able
rior
ound
ixel
case
a 15
first
n of
meter
S is
each
for:
table 3. adjustment results of analysis case with preprocessing (unit : m )
case 1 2 3 4
order position:1st position: 2nd position:1st position: 2nd
attitude:1st attitude:1st attitude:2nd attitude: 2nd
pre-pro 1B 1AP 1A 1B 1AP 1A 1B 1AP 1A 1B 1AP 1A
cessing
ox | 10.50| 7.18| 10.49| 9.59| 8.25, 12.19| 9.39| 13.02| 11.19| 10.93| 12.33| 12.88
preci-|oy | 11.71] 6.97, 10.17| 10.90, 8.01, 11.92| 10.57| 12.34, 10.94. 11.50. 11.55, 12.34
sion |oxy| 11.21] 7.08) 10.33) 10.27] 8.13| 12.06] 10.00] 12.68! 11.07] 11:17| 11.95] 12.61
Oz | 20.90| 12.02| 15.00| 18.89| 16.73| 17.43, 18.42, 26.42| 16.00| 19.89| 25.26! 21.23
tx | 18.00| 11.18| 13.31} 17.17| 12.24| 13.85/ 17.71| 20.70| 13.88] 16.98) 15.05) 16.96
accu- |uy | 14.26| 8.91| 7.44| 14.36| 11.17| 8.29| 9.12, 10.77, 9.00| 9.89| 25.57| 10.94
racy |mxy| 16.24| 10.11 10.79] 15.83] 11.72] 11.41] 14.02| 16.50| 11.70| 13.90| 20.98| 14.25
Uz § 17.00] 11,30] 6.42] 16.36] 18.14 7.01| 13.77) 20.51| 7.00) 20.51| 17.07} 12.07
image Vx |0.02810.0217|0.0153|0, 02700. 0261 (0.01770. 0262 (0.0224 |0.0177{0. 0224 |0. 0270/0. 0156
coordi | (mm)
nates
resi- e 0. 0260/0. 0181 |0. 0163 0. 02450. 0206 | 0. 0168 |0. 02380. 0174 |0. 0461 |0. 0370/0. 0370|0. 0149
dual mm
table 4. 3-D positioning
accuracy with preprocessing
E.| pre- geometric accuracy of image coordinate
precision (m) | check point (m) | residual(pixel)
0. |proce-
P. |ssing Oxy Oz Uxy Uz Vx Vy
15] 1B 110.00 | 18.42 | 14.02 13.77 1.048 0.952
42] 14P | 7.08 | 12.02 | 10.11 11.30 0. 761 0. 635
127 1A 110,33 | 15.00 | 10. 79 6.42 1.177 1.253
20
Ness)
Z.
ZB
ERROR (M)
Ji
7
7
7,
7
N
SIG XY SG Z MU XY
NNNNNNN IE
LEVEL 1B LEVEL 1AP LEVEL 1A
Fig. 5 3-D positioning accuracy with preprocessing
It was found through this analysis, that for photogrammteric
film products level 1AP is more suitable compared to 1B for
positioning due to its geometric stability. It was also found
that level 1AP showed better accuracy compared to 1A,
according to the difficulty in acquiring the image
coordinates. But when considering that level 1AP requires
analytical plotters, if methods to improve the
coordinates accuracy of level 1A is used, it can be said that
level 1A image data is the suitable data format for middle and
small scale map production as well as for establishment of
large topographic database.
image
375
3.3 Analysis of 3-D Positioning Accuracy According to
Adjustment Method
In this study,
positioning, AP's were introduced to polynomial such as
equation (3). Also,
detection and elimination of gross errors was developed and
to design a logical adjustment system for
a simultaneous adjustment method for
this was used in studying the effect of using different
adjustment methods. The significance testing which determines
Significant AP's was included in the developed simultaneous
adjustment method.
For comparing accuracies according to adjustment method, two
cases were studied. Case 1 is level 1AP with ground surveyed
controls and case 2 is level 1A with GCP from 1:50,000 maps.
For the case of level 1A with GCP from 1:50,000 map(case 2),
it was done to represent cases where mapping has to be carried
out for inaccessible area. (table 5)
table 5 accuracy of analytic case with adjustment analysis
method
ca| prepro- data aquisition method of G.C.P.
cessing
se| levels type 3-D coordinate|image coordinate
1 1AP photogrammetric| ground survey|analytic plotter
film
2 1A digital image | 1:50,000 map |eye measurement
3.3.1 Accuracy Analysis with AP's
For this analysis, a self calibration bundle adjustment with
18 AP's which includes signicance testing was used(Table 6,
Fig. 6).
For case 1, after adjustment with significant AP's and
geometric correction, the geometric precision and accuracy of
check points improved uniformly to less than 10 M.
