International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B1. Istanbul 2004
Katzenbeisser, R., 2003, About the calibration of lidar sensors,
ISPRS Workshop 2003, Dresden
Kraus, K. , 1997a, with contributions by J. Jansa und H. Kager
Photogrammetry - Advanced Methods and Applications.
Volume 2, Dümmler Verlag, Bonn, (English edition, translated
by P. Stewardson), 466 pages.
Kraus, K. , 1997b, Eine neue Methode zur Interpretation und
Filterung von Daten mit schiefer — Fehlerverteilung.
Österreichische Zeitschrift für Vermessung und Geoinformation
(VGL), 85. Jahrgang, Heft 1, 1997, S. 15 - 30.
(SCOP++, 2004) Interpolation, management, application and
visualization of digital terrain data. Product information:
http://www.ipf.tuwien.ac.at/products/produktin fo/scop/scop dt
m sheet.htm (accessed 15. Mar. 2004)
Wanninger, Lambert, 2003, Virtuelle GPS-Referenzstationen
für grófráumige Kinematische Anwendungen, zfv 3/2003
128.Jg.
APPENDIX
EXAMPLES TO DEMONSTRATE IMPROVEMENT
For checking the quality of geo-referenced laser scanner data,
the visualisation of height-differences between pairs of over
lapping laser scanner strips using (SCOP++, 2004) has proven
extremely valuable. Great areas of the surface can be checked at
one glance.
Palette This colour table was defined for
visualisations of Z-differences between two
; neighbouring ALS (Airborne Laser Scanning)
esent data strips using SCOP++. Positive differences
0.20 are coded in blue/ green and negative ones in
red/brown (associating heaven vegetation
0.12 ... carth).
Systematic Z-differences of open areas show
996 up as either greenish or brownish tones. Mere
XY-shifts are best seen on both sides of a
9005 pitched roof: one side will appear greenish, the
o.00s Other one brownish of same intensity. A shift
in the direction of the ridge will not show such
owe differences. Therefore another exposition of a
tilted surface is needed. When calibration and
0.12
geo-referencing both are perfect, the
difference-DSM should appear white, but due
to measuring and interpolation noise, the green
and yellow pastel colours should produce a
whitish random speckle.
BU uw om
Pit y el cuum $3)
raw data 0001 141 142 N p02
Figure 4a:
Figure 4b: adjusted data 0001 141 142 N p02
Figures 4 show the complete overlap area of strip 141(N>S)
and strip 142(S-? N). Asymmetries in the roofs disappear. The
B&W-printing obnubilates that to some degree. Height shows
the average approaching zero, with some few undulations left.
Figure 5a: raw data 0057 145 149 S p02
A P9
0057 145 149 S p02
Hn
*
data
i
adjusted
E LN
x?
Figure 5b:
Since the DSM in these examples is not filtered from
vegetation, woods and other noisy parts will appear as palette-
overflows.
560
