stanbul 2004
Combined
Point/Line
8.35
7.66
IRS-ID
mbined
rmed on the
ic IKONOS
)n, was used
t Cairo with
istribution
d projective
erent number
nts were the
accuracy of
image. RMS
.
27
control
points
1.79
1.65
IKONOS
ique with
im for point-
as performed.
ized in Table
ram.
nd combined
with only 5
trol points,
joints are also
the line-based
equivalent to
fewer control
ne techniques
id, thus,
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004
reducing tremendously the amount and expenses of fieldwork
needed.
RMS 7 Point- Line- Combined
URASC based based | Point/Line
X (m) 1.86 1.79 1.68 1.76
Y (m) 1.76 1.65 1.56 1.66
Table 11. RMS Values of 25 Check Points for IKONOS
Panchromatic Image Using Point, Line, and Combined
Point/Line Techniques
4. CONCLUSIONS AND RECOMENDATIONS
1- Linear image features are significant source of information
to facilitate image rectification process since they are
abundant in human-made infrastructure, and are amenable
to automatic feature extraction.
2- Projective equations can be based on corresponding point
features, corresponding straight lines features or
combination of point/line features.
3- Least squares adjustment techniques were developed for
point, line, and combined point/ line-based projective
equations.
4- The developed techniques were tested and analyzed by
performing several experiments using aerial photograph and
satellite imagery with various resolutions (LANDSAT7
panchromatic (15m), SPOT4 panchromatic (10m) and
mutlispectral (20m), IRS-ID panchromatic (5.8m), and
IKONOS panchromatic (1m)).
5- Results of the point-based projective equations technique
were compared to those obtained from two well-known
commercial software packages; Erdas Imagine version 8.5
and Intergraph (Z/I Imaging) I/RASC version 8.4, and they
proved equivalent. This confirmed the accuracy of the
developed program.
6- Experimental results of the line-based and combined
point/line-based projective equations techniques were
equivalent to those of the point-based projective equations.
However, only 5 control lines (for the line-based technique)
and 3 control line and 2 control points (for the combined
point/line-based technique) were used to achieve those
results, while an average of 25-30 control points were used
to achieve the same results (for the point-based technique).
This is very important to realize since it proves that, with
line-based or combined point/line-based techniques, the
number of control features can be reduced tremendously and
still yield the same accuracy figures. This will lead to the
reduction in the amount of field survey work and, thus, its
cost and duration.
7- Experimental results show the potential of using linear
features for the rectification of the time-dependent satellite
imagery. Study of specific geometry configuration of the
line features is important.
Research is continuing on the following:
l-Use of geometric constraints between various linear
features. This will provide substantial information in
support of photogrammetic restitution and image
rectification process.
2- Develop affine and polynomial equations (with different
orders) using straight lines features.
3- Develop mathematical models (affine, polynomial, and
projective) using linear features in general and not only
straight lines, such as conic sections (e.g., a circle on the
ground and its projection as an ellipse on the image).
4- Development of line-based and combined point/line-based
for other photogrammetic equations/conditions and
mathematical models for image rectification.
5- Experimentation to study the effects of various
configurations of control lines (vertical, horizontal, parallel,
etc), and combination of control points and lines on the
performance of the developed technique.
5. ACKNOWLEDGEMENTS
The authors would like to express their gratitude to Prof. Dr.
Mona ElKady, Chair of the National Water Research Center,
for her support and encouragement. Thanks are also due, Prof.
Dr. Rafeek AbdelBari, the director of the Survey Research
[Institute for granting the use of the software and the various
satellite images in the experiments.
6. REFERENCES
[Barakat, 1997] Barakat, H.F. 1997. Photogrammetric
Analysis of Image Invariance for Image Transfer and Object
Reconstruction, Ph.D. thesis, School of Civil Engineering,
Purdue University, West Lafayette, Indiana, USA.
[Barakat et al, 1995] Barakat, H. F., Weerawong, K., and
Mikhail, E. M., 1995. Comparison between Invariance and
Photogrammetry for Image and Object Transfer, Conference
on Integrating Photogrammetric Techniques with Scene
Analysis and Machine Vision, SPIE Proceedings: Vol. 2486,
pp. 13-24, Orlando, Florida, USA.
[Mulawa and Mikhail, 1988] Mulawa, D., and Mikhail, E. M.,
1988. Photogrammetric Treatment of Linear Features, ISP
Congress, Commission III, Vol. 10, pp. 383-393, Kyoto,
Japan.
[Shaker et al, 2002] Shaker, A., Barakat, H. F., and Shi, W.,
2002. Accuracy Assessment for Different Cases of IKONOS
Imagery using Two-Dimensional Models, Proceedings of the
3" International Symposium: Remote Sensing of Urban Areas,
pp. 139-146, Istanbul, Turkey.
[Weerawong, 1995] Weerawong, K., 1995. Exploitation of
Linear Features for Object Reconstruction in Digital
Photogrammetric Systems, Ph.D. Thesis, School of Civil
Engineering, Purdue University, West Lafayette, Indiana,
USA.
