Jsing Match-
tched feature
nage and the
ipolar condi-
matched the
joints on the
In this exam-
ers 1 and 2 in
points along
1e 1-2 in the
esolved using
ween feature
e well known
jest solutions
paralell lines.
re more than
oint Approxi-
tial matching
e features. In
y tenth edge
ature corners
in the left im-
5 with the rel-
e right image.
for line seg-
ear-horizontal
ire 8.
icy, Sub-pixel
strically Con-
| described by
Figure 8: Zoomed Sub-images of Poorly Defined Epipolar
Line Intersections
7 CONCLUSION
We have presented an image matching scheme that incorpo-
rates both feature- and area based matching techniques. A
novel, two-stage feature matching algorithm was presented.
The two stages of the matching scheme take into account
firstly the local structure of a feature, and secondly the spa-
tial relation between a feature and its neighbours. Combining
these two aspects improves on matching schemes that em-
ploy only one or the other. We have also shown how feature
matching can be used to obtain initial estimates to the rel-
ative orientation, effectively enabling us to subsequently re-
fine this relative orientation using a combination of feature-
and area based matching and the resultant epipolar geom-
etry. This technique makes it possible to carry out a fully
automated relative orientation without any prior knowledge
of orientation parameter estimates.
Improvements can be made to this matching scheme, most
notably in the handling of ARCs and images with large dis-
parities. Extending this matching scheme to a multi-photo
and multi-resolution environment will also help in improving
the results obtained.
ACKNOWLEDGEMENTS
The authors would like to extend their gratitude to the Foun-
dation for Research Development (FRD) and the University
of Cape Town (UCT) for their invaluable support for this
research. :
REFERENCES
[1] E. Baltsavias. Multiphoto Geometrically Constrained
[2]
[3]
[4]
[5]
[6]
Matching. PhD thesis, Institute of Geodesy and Pho-
togrammetry, 1991. Mitteilungen nr.49.
J. Canny. "A Computational Approach to Edge Detec-
tion". IEEE Transactions on Pattern Analysis and Ma-
chine Intelligence, PAMI-8(6):679-698, November 1986.
S. Cochran and G. Medioni. “3-D Surface Description
from Binocular Stereo”. /EEE Transactions on Pat-
tern Analysis and Machine Intelligence, 14(10):981—994,
1992.
G. Cox and G. de Jager. "A New Point Pattern Recogni-
tion Technique". Technical report, Department of Elec-
trical and Electronic Engineering, UCT, 1994.
G. Cox, G. de Jager, and B. Warner. "A New Method of
Rotation, Scale and Translation Invariant Point Pattern
Matching Applied to the Target Acquisition and Guiding
of an Automatic Telescope”. 2nd South African Work-
shop on Pattern Recognition, November 1991.
H. Freeman and L. Davis. “A Corner-Finding Algorithm
for Chain-Coded Curves". IEEE Transactions on Com-
puters, pages 297-303, March 1977.
709
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996
[7] A. Gruen. “Adaptive Least Squares Correlation - A Pow-
erful Image Matching Technique”. Presented Paper to
the ACSM-ASP Convention Washington D.C., 42:97-
112, March 1985.
A. Gruen and E. Baltsavias. “High-Precision Image
Matching for Digital Terrain Model Generation”. Pho-
togrammetria, 42:97-112, 1987.
A. Gruen and D. Stallmann. “High Accuracy Dimen-
sional Measurement Using Non-Targeted Object Fea-
tures”. International Archives of Photogrammetry and
Remote Sensing, 29(B5):694—700, 1992.
R. Haralick and L. Shapiro. Computer and Robot Vision,
volume 1-2. Addison Wesley, 1992.
O. Hellwich and W. Faig. "Graph-Based Matching
of Stereo Image Features”. International Archives of
Photogrammetry and Remote Sensing, 29(B3):307-317,
1992.
O. Hellwich and W. Faig. "Graph-Based Feature
Matching Using Descriptive and Relational Parame-
ters". Photogrammetric Engineering and Remote Sens-
ing, 60(4):443-450, April 1994.
R. Horaud and T. Skordas. "Stereo Correspondence
Through Feature Grouping and Maximal Cliques". /EEE
Transactions on Pattern Analysis and Machine Intelli-
gence, 11(11):1168-1180, November 1989.
A. Oppenheim and R. Schafer. Discrete - Time Signal
Processing. Prentice-Hall, 1989.
A. Rosenfeld and A. Kak. Digital Picture Processing,
volume 1-2 of Computer Science and Applied Mathe-
matics. Academic Press, Orlando, second edition, 1982.
T. Schenk. "Algorithms and Software Concepts for Dig-
ital Photogrammetric Workstations". In D. of Geode-
tic Science and Surveying, editors, Technical Notes in
Photogrammetry, July 1992.
T. Schenk, J. Li, and C. Toth. "Towards an Autonomous
System for Orienting Digital Stereopairs". Photogram-
metric Engineering and Remote Sensing, 57(8):1057-
1064, August 1991.
F. Stremler. Introduction to Communication Systems.
Addison-Wesley, second edition, 1982.
A. Tabatabai and O. Mitchell. “Edge Location to Sub-
pixel Values in Digital Imagery". [EEE Transactions
on Pattern Analysis and Machine Intelligence, PAMI-
6(2):188-201, March 1984.
N. van der Merwe and H. Rüther. "Image Matching
Through a Combination of Feature- and Area Based
Matching". International Archives of Photogrammetry
and Remote Sensing, 30(5):407-413, March 1994.
