International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004 ni 
ntern 
  
  
3.2 The control schemes for Ikonos block adjustment and 336 check points for When 
examining the production have been collected. All the GCPs 
It is very hard to do the field survey in the area. Except for the and the check points were surveyed by Leica GPS 530 with een 
factor of the high altitude and the inconvenience traffic, the static or rapid static survey, and its accuracy is x5cm. Because 
visual condition along the surface of the ground is very worse. of the craggedness of the mountain, the GCPs are distributed 34 1 
The GPS receivers were adopted in the field survey so as to equably in 3 stereo images among the valley. There are about There 
obtain the control points and check points of Ikonos images. In 20 GCPs in each stereo images(see Fig.1). di 
ë 
order to test the accuracy of the Ikonos stereo image, 58 GCPs 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
No. 
1 
2 
3 
4 
5 
6 
7 
8 
9 | 
10 A 
11 
12 | À 
13 
14 
15 
16 
17 
18 
19 
20 
Note: S 
Fig. 1 The project area and the distribution of the GCPs 
3.3 The Ikonos accuracy without control It can be seen obviously from table 2 that there are bias and TI 
a : Mn N ; 1e blo 
: ; drift in Ikonos images and they are different from image to e 
When there is no GCPs, the coordinate and its errors of the image. 
image space and the ground space are calculated respectively €  Theerrors in ground space without GCPs con 
with the RPC provided by the vender. The errors between the ground coordinates calculated from e 
€  Theerrors in image space without GCPs RPC model and its real ones are listed in table 3. the | 
The image coordinates calculated from the RPC model are dist 
compared with its measured coordinates, its errors are listed in . 
table 2. GCI 
; e 
the | 
03 05 07 09 12 aon long 
lage ID I s l S I S 1 s l S I s heig 
Max. error | 9.471 | -5.618 | 4.477 | -4.723 9262 | -3.975 5.410 | -3.143 7.847 | -5.145 4.410 | -2.558 e 
Mere i. a4] „7238 1.2.2951 6341| 7.0381 -6.107 | 3519| -5.285 | $919 | -7.017 | 2.606 } -4.159 exis 
ee cues | S207]. 6477]. 3391] 56471 8.333 | -5015| 45861 -4.180 | 67359| -6291 | 2476 |. 3.57 accu 
RMS 0.826 0.558 0.749 0.515 0.614 | 0.662 0.648 0.709 0.654 0.525 0.595 0.481 in E 
Table 2 The errors of image space without GCPs(pixel) NDS 
3.5 The 
Stereo 1(03/05) 11(07/09) III(12/13) 
Model X Y H X Y H X X H The acci 
Max. error -6 188 | -5247] 8.044 | -5.156 | -5.383 | 7.764 | -7.164 | -3.904 7.709 with Iko 
Min. error -8.882 | -8.129 | 3.990 | -7.960 | -7.649 | 3.600 | -8.975 | -6.103 5.330 plane an 
Ave. error 772 | 6399 | 7.401 | -6.980 | -6.429 | 6.253 | -8.278 | -4.983 6.466 the DLG 
RMS 0.827 0.976 1.071 0.802 | 0.730 | 0.964 0.535 0.695 | 0.622 
  
  
  
  
  
  
Table 3 The errors in ground space without GCPs(meters) 
1046