'stanbul 2004
th images in
ierarchy and
mages on the
le pairs,
res are tied
| levels using
Tay 5
(OW AE
poss
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004
- proceed with features verification using epipolar geometry
restrictions,
- estimate orientation parameters using the coarse-to-fine
approach to increase the accuracy of result.
The approach under consideration was implemented and tested
in digital photogrammetric system Z Space (Blokhinov, Y.,
Sibiryakov, A., Skryabin, S., 2000). Two examples of the tests
are shown in Figure 11 and 12. Accuracies for standard
deviation of the image coordinates are 0.36 and 0.27 pixels
respectively.
Figure 11. Accepted conjugate points, automatic relative
orientation of aeroborn images
Figure 12. Accepted conjugate points, automatic relative
orientation of spaceborn images TK-350
Now it’s up to user to decide, whether the approach concerned
meets the requirements of practice.
4. CONCLUSION
In the presented paper the universal approach to image
matching for various types of imagery is proposed. At first,
point features are extracted and filtered, using the original
concept of “informativity size”. The conjugate pairs of features
are tied according to their vicinity in parameter space and
filtered using the relational distance graph, both measures being
invariant to image rotation. In what follows the conjugate points
found can be used in different ways according to final goal of
specific work.
In this study the approach was applied to the tasks of automatic
mosaic creation from video camera sequence frames and
automatic relative orientation of photographic camera images.
In all cases the approach concerned shows high reliability and
computational efficiency.
References
Blokhinov, Y., Sibiryakov, A., Skryabin, S., 2000. Z Space -
digital photogrammetric system for Russian TK-350 images.
The International Archives of the Photogrammetry and Remote
Sensing, Amsterdam, Vol. XXXIII , Part B2, Com. II, 81-90.
Heipke, C., 1996. Automation of interior, relative and absolute
orientation. The International Archives of the Photogrammetry
and Remote Sensing, Vienna, Vol. XXXI, Part B3, pp. 297-311.
Henricsson, O., 1996. Analysis of Image Structure using Color
Attributes and Similarity Relations. Dr. sc. techn., Institute of
Geodesy and Photogrammetry Swiss Federal Institute of
Technology, Zurich.
Hu, M.K., 1962. Visual pattern recognition by moment
invariants. IRE Trans. Information Theory, Vol IT-8, pp. 179 —
187.
Tang, L., Heipke, C., 1993. An approach for automatic relative
orientation. Optical 3D Measurement Techniques II:
Applications and Inspection Quality Control, and Robotics, pp
347-354.
Woozug, C., 1996. Relational Matching for Automatic
Orientation. The International Archives of the Photogrammetry
and Remote Sensing, Vienna, Vol. XXXI, Part B3, pp. 111-119.
