International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004 Intei 
— 
  
  
  
  
  
  
  
  
  
  
  
    
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
   
     
     
   
  
Hopes]: 23s6- | 3549:] 3332. | 2246 | 1230 | 3.2 
a 12 246 340 490 2 S 
C, = 2 =3.554 m. = 0,187 0.246 0,340 1.499 1,002 0,512 For 
n 13 | 0,834 1,284 1,852 0364 | 0295 | 0293 m 
14 | 476: | 4663 |.-5627 | 1790 | 2787 | 4023 full 
7. = s — 4 375m. 15 | 1,034 | 0,920 1.022 | 2,794 | 2418 | 2,609 i 
n 16 1,325 1,742 2,415 3,917 4,084 5,134 
DO 171 0997 | 0994 | 1214 | 3,788 | 3,028 | 2,928 
oc, to, 18 2 1 : 
c. = Zo -4712m e |o 11s |. 1405 ] 3517 | 2,058.] 1.758 
2 19 | 5024 | 4998 | 6100 | 1222 | 1,586 | 2.91 
20 4 a kan & 
: ; : . 14 4,467 3,523 483 5.8 5.34 
Error band illustrates the accuracy for the lines in much more 51 S a6 > 3 7.1% d 49 | 
understandable and interpretable way. In order to draw these = 4,258 3,020 2,163 3,077 3,408 4,424 
error bands for all the lines in sampling group, proportions are 22 | 0,673 1,249 1,856 1,190 1,063 1,188 T, 
selected as %, + and ^4. The deviations on these portions of the 23. 4.634 3.465 2.922 2.926 4.702 6.834 ; 
sampling lines are calculated as in Table 3. 24 2752 5.865 4 067 0.850 6573 p ADD 
= yf J& fy 2 3,U0 ;9- 39 3 33 pp 
[m] t m 25 | 42699 | 6607 | 2071 1.220 | 2548.1 3984 samy 
; Deviations £y Deviations £, 26 : : : = zi 
/ ; : : : € „895 5,032 3.965 3393 ‚0 RS ee 
# Portion % | Portion “| Portion % | Portion '4| Portion 7| Portion Ya 57 6,809 03 9 11,393 10011 10.983 Lanc 
= 288 7,189 8,722 14,1( 4,072 2 ; 
| | 1953 | 3742 | 5586 | 6,180 | 8,667 | 12231 = LES ou I.S Jol | 14.072 4 17.212 | Ty 
os 8, 8,232 9,702 ,143 215 3 
2 | osas | 0,764 | 0,860 | 2,167 | 1,974 | 2246 e: JS | Ss 20 04 0.215 1 0.106 | 1.1. 
= S78 3,73 1,928 2,39 2,538 3 
3 | 2014 | 1000 | 1.130 | 0382 | 0268 | 0.182 = son 9 s 35s 538 1 3217 41 
5 3,415 2,85 2,95 1,36 ‚828 2 
4 | 1438 | 2861 | 4290 | 1,444 | 0,965 | 0,495 meld I 281 22) 360 | 182% 1 2501 +41 
= Table 3. Deviation on predefined portions along sampling lines 44: 
T 0,601 0,400 0,200 1,236 0.824 0,412 L 
6 0,609 1,042 1,530 1,240 0,831 0,431 Lt 
7 | 9421 0.342 0,339 0,130 0.089 0,053 Using the deviations given in Table 3, all error ellipses are 1.1. 
8 à; de > e drawn along the sampling lines in order to show the error bands 1.4. 
9,500 0,360 9,362 3467 2201 S19 for all the sampling lines as in Figure 3. 12. 
9 | 2521 | 2,014 | 1946 | 2493 | 1,672 | 0,880 E 
2. 
10 | 0763 | 0888 | 1,179 | 0697 | 1.077 | 1,554 = 
12. 
12. 
1.2. 
12. 
12. 
12. 
12.1. 
1.2.2 
1.2.2 
1.2.2 
Figure 3. Error ellipses(red) and error band (blue) of a line (black) 1.2.3 
iki Ep su : 1.2.2 
A generated accuracy criterion for areas is given in Tastan | 4 edges 
(1999) as follows. FT 
-—ib-9 edges |. ld ieg 
S edoos |: After 
360 edges TE usin 
o, —0. |2FSinl —— | | 8 
Fr er | a 1 10 edges ERT are idi 
| —— 20 edges, ri, 
where F is the value of the area and n is number of edges. Using . —@— 50 edges For al 
this formula, geometric accuracy of areas can be given as a T proces 
graph. i m mee 
As it can be seen from the formula, accuracy of an area depends 10 ema 
on the area and number of edges of itself. Figure 4 shows | 0 Contro 
dependencies of accuracy of area with its area and its number of us ae Slag proces 
d Area (ha) have t 
edges. | Errors 
Figure 4 Geometric accuracy changes of areas from different sampli 
edge numbers and areas 
1028