ul 2004 
mation 
m each 
t0.1m 
y of the 
as the 
ements 
without 
lent for 
/INS is 
equests 
terrain 
lanned 
770 km 
rent of 
iter to 
based 
1ced to 
“GPS, 
) km. 
CGPS 
ational 
(d in a 
t turns 
not to 
ational 
m. 
ome a 
hnical 
  
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B1. Istanbul 2004 
  
The calibration field used to be a small area covered by two 
strips with six to eight images each, where side and forward 
overlap were set to 60 percent. This requirement was relaxed 
by IGI to 60 percent forward and 30 percent side lap with the 
latest software solution, so any part of a usual survey flight 
can be used. 
The time spend on a flight line should not exceed 20 min. in 
order to restrict the drift for heading to the specified accuracy 
of 0.01 grades. If an integrated sensor orientation is planed, 
this restriction does not apply, because the heading is of less 
influence. 
2.1.3 The survey flight has to start with an initialising phase 
of some minutes for the system while still grounded which is 
needed for the real time solution during flight rather than the 
post processing. 
Shortly before arrival at the project area, a s-turn had to be 
flown to activate the inertial system. This requirement was 
also relaxed by IGI lately with the latest software solution. 
At the end of the strips, left hand side and right hand side 
turns should alternate.. This is to avoid possible cumulating 
of errors. 
2.1.4 Ground control has to be established within the 
calibration field. Usually natural points are chosen (man 
holes, corner of building, e.g.) and co-ordinates provided by 
local survey. In some cases, co-ordinates are taken from 
cadastral information or even a map. The accuracy needed 
depends of course on the image scale. 
2.1.5 Computation of the projection centers can be done 
using Aerobic'M, where the GrafNav DGPS software 
package of Waypoint Consulting Inc. is delivered with. Other 
packages can be used as well. 
2.1.6 The triangulation of the calibration field are done at 
Hansa Luftbild German Air Surveys digitally, utilising the 
GPS positions as additional observations. The RMS of the 
projection centers is set to 0.01 m, in order to keep them 
nearly fixed. 
2.1.7 Exterior orientation values for all images arc 
computed finally. The difference between the exterior 
orientation values for c, « and x determined by GPS/INS and 
the triangulation for the images of the calibration field serve 
as calibration values. This is applied to all other angular 
measurements of the project. 
3. Examples 
3.1 Project Hessen 
The federal state of Germany Hessen (21115 qkm) had to be 
covered with CIR-images scale 1: 24.000 within a short 
period of time in early summer 2003. About 2800 images (60 
% forward overlap) have been taken with 12 flights within 8 
days, using 3 planes, and three cameras (two at the time), 
equipped with GPS/INS. The IMU's had to be changed 
between different cameras. 
In order full fill the very tight time schedule to produce ortho 
photos, it was decided to scan only every second image and 
‘ 
use the results of calibrated GPS/INS directly without overall 
triangulation. Distributed over the area, three calibration 
fields have been established. 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
x N 
NE = 
= S- S5 = té sn 
= 3 Il = X „= 
9| 2 alt = = = T 
> = - = o x — o = 
Fl ES Vy rule fg +5 = | 22 
Dill Du | Z = | & a. = 
S en defo TG orton on | wo 
DU € E e un = <t Ci QN 
e © oO e + © = e 
+ = e ol = c e S S 
—“|—- als 15 els cis 
[- en 
. oo — > e > 
CA -€-— 9$ | | = = | A 
= e1 © [o] e un c c © 
HS SEAN SS 
—|—|— e| o ; > c c = 
~ > 
> > > — + 
ple > 17 V0 on oy | + 
m OS Ce S pe 
: C 2 e = S 
un <3 . - 
ss SE las iS (ale eis 
> > 
. on oo oo wv e 
p 2 E. 00 = ON OÙ en uv 
© ©) p- = + c © © 
S| =| A = — c © © 
—_ | — = af Ss 7 c c c c 
oo uv 
e o a oc > 
e A SH Fed a | 
EIS Solo |B EB | SE 
AE do VS SS S |S 
oo > 
: + oo E e - 
Neo) = N en x Ss en = > 
elio Ro e^ = e e e 
po PA o aue ES e Pe 
ed ri rm — e 7 2 c e 
) ON 
= e1 [€T 12 re 
> ca — = C1 ce cn oo 
e m Q e > > = = 
ca - aD el c c © = = 
— — 
— -—| — -l c 1 1 > = e 
20 un C 
S C1 + — I 
Eie m > |% |2 |: « ce] 
om t real a. 8 S | = 
S19 Bo Ve 38 BB 
cM — — €-—Í| c 1 T C c = 
en 
ES e S e e © > 
A| Ye) nn cn = C1 un 
© esi pec] en co = C c — 
S| FE C = = S & e 
—_ | — = — © ; ETE e = = 
= 0 n = en > wu 
pU eri e (o = + C4 on 
© — © C1 > c <2 © = 
S | X ee hme es 
CN €-—| — Tic 1 1 e © © 
= c 3 ON uy — | 
JS Er. ~ eS > CN + oo 
eJ. e c un e S = 
els x e = e S = e 
ma = «| © ; c > c e 
o en v co % c uv 
eu e na s a | = 
eee T e = © e e e 
Lu = = S S S S 
EIE e c 7 c e c e 
o s | |s | - n|e 
ei s & |m-«s mS + | > 
eiq me 2 o 2 = & = 
SS e = = S S e 
a =] = Qu o 1 > ec c c 
  
  
  
  
  
  
  
  
  
  
  
Table 1: results of GPS/INS for project Hessen