' positions 
Prence of 0.6 
«factor of the 
which results I 
■o the north 
ng direction, 
ues are in the 
duced below 
identified and 
nised with the 
in the digital j 
eloped which ■ 
in all images, 
: to determine j 
r of cameras 
e orientation, 
m system into 
ng parameters 
an. The final 
based on the I 
of exposure, a 
if the relative I 
mbyKlemm 
including the 
ic measuring 
a geographic 
ns and to add 
Atfj /kis>/0010/t$Offto/b*b \ 01/oifOJ/(H)O/O*'»0 im 
¡13 j 6 
r 'J ,St 1193 
lAimehm®« 
SthilUpi Stirtfte BotJ-Anlog«n 
AHgcmein Balnwnlage 
ncm 
C£j f 
Salzburg 
Figure 7: Object Measurement with KiSS and attributes for 
geographic information systems 
4 EFFICIENCY AND ACCURACY Of KiSS 
In a pilot project, the efficiency of the system could be 
demonstarted. The aim of the project was to determine the traverse 
of the highways and to record the environment, the junctions, traffic 
lights, bicycle tracks and noise protection walls in three rural 
districts. The measured highways are shown in figure 8. 
For this aim 660 km of highways had to be measured in both 
directions. The area has an expansion of 50 km times 80 km. The 
measurements were carried out in three campaigns of 16 days in 
total and two additional days for the evaluation of the 
transformation parameters from the GPS system to the local 
coordinate frame. One measuring session was fixed to about one 
hour of driving. The satellite shadowing caused by villages and 
forests were at the most 5 minutes. The rate of failure caused by 
strong solar radiation or rain drops on the lenses or hardware 
problems were less than 5 %. These sections had to be remeasured. 
The object measurement for the determination of intersections and 
the environment took less than 40 man-days. 
The accuracy of the system was checked with 25 intersections of 
different routes. The standard deviation of the two dimensional 
position was about 0.6 m. This uncertainty includes the errors of all 
sensor calibrations, including the photogrammetric calibration of 
the CCD sensors, the errors in the determination of the trajectory as 
well as the errors of the object measuring. 
Figure 8: Streets and intersections 
2-2-5