flight time of about one hour. 77 photographs out of a
total of 168 were selected for the block adjustment. The
resulting parameters of exterior orientation determined by
photogrammetric means had standard deviations of
about 3 cm in each of the coordinates and 7 arcseconds
in each of the orientation parameters. The discrepancies
between these parameters and those determined by
GPS/INS were much larger. The root-mean-square (rms)
discrepancies were 15 cm in horizontal position, 20 cm in
vertical position, 1 arc minute in azimuth, and two arc
minutes in roll and pitch. When these parameters directly
determined by GPS/INS were used for image
georeferencing without the use of ground control, the rms
discrepancies on control points were 0.3 m horizontally
and 0.5 m vertically. This is sufficient for most mapping
and resource applications. For more details, see Skaloud
et al (1996).
These results are quite unexpected and need further
analysis. The error pattern for roll and azimuth seems to
indicate a major influence of errors due to aircraft
dynamics, while the pitch error shows a linear drift with
time. The fact that, in contradiction to theory, the azimuth
performance is better than that in roll and pitch, might be
due to the frequent 360° turns, necessary to return to the
test area. However, these results should be considered
preliminary until more detailed investigations have been
done. They show that it is very difficult to reach an
attitude noise level of 20 - 30 arcseconds over an
extended period of time using current hardware.
5. CONCLUSIONS
Georeferencing of airborne imaging sensors has the
potential of considerably extending current
photogrammetric applications and greatly simplifying the
use of digital imaging sensors. It also adds flexibility to
the use of current high-precision aerial cameras and
considerably reduces the need for accurate ground
control.
The integration of inertial and GPS satellite techniques
currently offers the best potential for implementing
georeferencing systems at different levels of accuracy
and for combining them with existing and future airborne
imaging sensors. Major advantages are the high data
rates of the inertial measuring unit, compactness which
allows direct mounting on the sensor head, and uniform
high accuracy due to continuous GPS updating.
Theoretical studies and test results, analyzed in this
paper, indicate that current systems are capable of
georeferencing airborne imaging sensors for mapping
and resource applications with an accuracy of about 0.5
m (rms). Higher accuracy can be achieved if the long-
term stability of the attitude sensors can be improved to a
noise level of about 20".
6. ACKNOWLEDGEMENTS
Financial support for this research was obtained through
a grant of the Natural Science and Engineering Research
Council of Canada. Messr. J. Skaloud, Q.J. Zhang, and
Y. Li, graduate students at the Department of Geomatics
72
Engineering, and Dr. M. Wei are thanked for their
contributions to this research.
7. REFERENCES
Ackermann, F. (1994): On the Status and Accuracy
Performance of GPS Photogrammetry, Proceedings
'Mapping and Remote Sensing Tools for the 21*
Century', pp. 80-90, Washington, D.C. August 26-29.
Cannon, M.E., and G. Lachapelle (1992): Analysis of a
High Performance C/A Code GPS Receiver in Kinematic
Mode. Navigation, Journal of the U.S. Institute of
Navigation, Vol. 39, pp. 285-299.
El-Mowafy, A. (1994): Kinematic Attitude Determination
form GPS. Ph.D. thesis, UCGE Report No. 20074,
Department of Geomatics Engineering, The University of
Calgary, Calgary, Alberta.
Hothem, L., K. Craun, M. Marsella (1994): Operational
Experience with Airborne GPS Controlled Photography at
the US Geological Survey. Proceedings 'Mapping and
Remote Sensing Tools for the 21% Century’, pp. 51-63,
Washington, D.C. August 26-29.
Lachapelle, G., R. Klukas, D. Robert, W. Qiu (1994):
One-Meter Level Kinematic Point Positioning Using
Precise Orbits and Satellite Clock Corrections.
Proceedings of ION GPS-94, pp. 1435-1433, Salt Lake
City, Utah, September 20-23.
Lucas, J.R. (1994): Covariance Propagation in Kinematic
GPS Photogrammetry. Proceedings ‘Mapping and
Remote Sensing Tools for the 21% Century’, pp. 91-101,
Washington, D.C. August 26-29.
Schwarz, K.P., M.E. Cannon, R.V.C. Wong (1989): A
Comparison of GPS Kinematic Models for the
Determination of Position and Velocity along a
Trajectory. Manuscripta Geodetica, 14 (2), pp. 345-353.
Schwarz, K.P., M.A. Chapmann, M.E. Cannon, P. Gong
(1993): An Integrated INS/GPS Approach to the
Georeferencing of Remotely Sensed Data.
Photogrammetric Engineering and Remote Sensing, 59
(11), pp. 1667-1674.
Schwarz, K.P., M.A. Chapmann, M.E. Cannon, P. Gong,
D. Cosandier (1994) A Precise Positioning/Attitude
System in Support of Airborne Remote Sensing.
Proceedings of the GIS/ISPRS Conference, Ottawa,
Canada.
Schwarz, K.P. and G.S. Zhang (1994): Development and
Testing of a Low-Cost Integrated GPS/INS. Proceedings
of ION GPS-94, pp. 1137-1144, Salt Lake City, Utah,
September 20-23.
Schwarz, K.P. and M. Wei (1994): Modelling INS/GPS for
Attitude and Gravity Applications. Proceedings High
Precision Navigation 95, pp. 200-218, Stuttgart, April 3-6.
Dümmler Verlag, Bonn.
Schwarz, K.P. (1995) Integrated Airborne Navigation
Systems for Photogrammetry. Proc. Photogrammetric
Week'95, Stuttgart, May ..., 1995, pp. 139-153.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B6. Vienna 1996