2004 
  
Ch of 
avia, 
nine 
erior 
show 
1 are 
ding 
id 50 
s are 
AAT 
| one 
"TOSS 
O is 
;pec- 
erest 
rect 
erial 
| this 
'hich 
tion: 
rs in 
evia- 
lirect 
2.2), 
ite a 
they 
s are 
1 the 
taset 
sfore 
bally 
nish- 
for 
than 
ong- 
and 
1g60 
, but 
1 the 
with 
sults 
m in 
y of 
gure 
enta- 
es a 
1Xes, 
le in 
they 
light 
esn't 
etric 
ome 
| are 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004 
  
6. CONCLUSIONS 
The results of a vast and rigorous test on quality of direct 
georeferencing on photogrammetry have been shown. They are 
interesting and encouraging. Some issues have also emerged 
and in particular the existence of a limited systematic error in 
height whose cause is probably a residual miscalibration. 
Further developments will focus strategies for limiting these 
effects: more frequent calibrations or local calibration on a very 
small site. 
7. BIBLIOGRAPHY 
Casella, V., Galetto, R., Surace, L., Ferretti, L., Banchini, G., 
Cavalli, A., 2001. Esperienze di fotogrammetria supportate da 
GPS/INS. Bollettino SIFET, 4, pp. 35-49, ISSN 0392-4424, 
Parma. 
Casella, V., Franzini M., 2003. Definition of a methodology for 
local reduction of parallaxes in directly oriented images. 
Proceedings of "/SPRS International Workshop", WG 1/5, 
September 2003, Castelldefels, Spain. 
Colomina, I., 2002. Modern sensor orientation technologies and 
procedures. Test Report and Workshop Proceedings. Official 
OEEPE Publication n?43, pp. 59-70. 
8. ACKNOWLEDGEMENTS 
The research presented in this paper was carried out within the 
frame of the National Research Project entitled /ntegrated 
inertial positioning systems in aerial Photogrammetry, co- 
funded by the Italian Ministry of the University for the year 
2002, and chaired by prot. Galetto of the University of Pavia. 
APPENDIX A: GRAPHICAL ANALYSIS OF RESULTS 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
Delta E - Whole Delta N - Whole Delta h - Whole 
0,70 1 0.60 1.20 A 
-*- Pavia ~~ Pavia ; — Pavia 
0.60 5— Perugia 0.50 75 Perugia 1.00 - Perugia 
—- Como — Como —&— Como 
a 0.50 Palermo 4 Palermo zz Palermo 
$ # — Parma = 0.40 Parma £ 080 os 
9 0.40 Vercelli 2 Vercelli 2 Vercelli 
E E E 
— . = (0,30 = 0.60 - 
1 S = = 
% 0,30 RC NC z 2 
5 ”" Om € 0.20 : Z 0.40 on 
0,20 ; = : = a Se 
NS — e 
0,10 0,10 0.20 
0.00 0.00 0,00 
APP AAT AT APP AAT AT APP AAT AT 
Figure 1. Analysis of accuracy: RMSE of the residuals for the Whole dataset, as a function of the EO 
Delta E - Along 60 Delta N - Along 60 Delta h - Along 60 
0.70 0.70 0.70 "x 
*— Pavia -*- Pavia An E 
0,60 9— Perugia 0,60 $— Perugia 0.60 x R 
—# Como —&- Como NES 
z 0,50 Palermo = 0,50 Palermo = 0,50 
S : Parma S Parma S 
E 0.40 C Vercelli E 0.40 s ! Vercelli E 0.40 
2030 & 930 B 030. e Pavia 
= E. a = x = == Perugia 
& 0,20 En TE - es X 0,20 M et ® 0,20 ia Como 
Palermo 
0,10 0,10 0.10 “Parma 
Vercelli 
0.00 0.00 0,00 
APP AAT AT APP AAT AT APP AAT AT 
Figure 2. Analysis of accuracy: RMSE of the residuals for the Along60 dataset, as a function of the EO 
Delta E - Across Delta N - Across Delta h - Across 
0,70 0.70 1.20 
+= Pavia *— Pavia += Pavia 
0.60 -9- Perugia 0,60 Perugia 1.00 Perugia 
a= Como ——- Como —— Como 
= 0.30 Palermo ~ 0,50 Palermo 7 Palermo 
e + Parma £ * Parma © 0,80 Parma 
& 040 : Vercelli 9$ 040 Vercelli = Vercelli 
E NN £ ; = 0.60 
& 030 NON & 0.30 S A \ 
= NN = ng = 040 éme 
& 0.20 . X 0.20 i i & a 
eu = : wr 0.20 
0.10 — 0,10 I ; 
0,00 0.00 0.00 
APP AAT AT APP AAT AT APP AAT AT 
  
  
  
  
  
Figure 3. Analysis of accuracy: RMSE of the residuals for the Across dataset, as a function of the EO 
883