is used by us to generate initial approximate
orientation parameters along orbital paths. We
conclude that fewer than 10 ground control points
are quite enough for space resection (Chen and
Lee,1990). It is interesting to note that the re-
section procedure is equally applicable to a sin-
gle SPOT scene or to multiple scenes in stere-
oscopy. As far as improvements on additional
self-calibrating parameters are concerned more
research on our part is necessary.
One of the purposes in studying our integrated
model for image matching and 3D positioning
is to test its applicability to spaceborne stereo
imagery. We know there is an advantage in theo-
retical analysis on error propagation; it is a wor-
thy trade-off when thinking about rather heavy
computational loads. Another advantage inher-
ent in the integrated approach is automatic gen-
eration of orthographic (sub-)images along with
that of a dense digital elevation model. Cur-
rently, we must impose weights in form of a di-
agonal matrix on unknown height parameters of
DEMs. This arises in part from the fact there
exists locally no or little image contrast in SPOT
sub-images. Therefore, our on-going research ef-
forts are led to
- determine in image preprocessing weighting
functions that allow for gray-value contrast,
textures or features in (sub-)images;
- design regional radiometric parameters to re-
place two global parameters ro, r1 in Eq.(3b);
- discover methods of interpolation more realis-
tic than those by non-differentiable, contin-
uous piecewise linear or patchwise bilinear
modeling;
- as a long-term goal correct radiometrically for
atmospheric effects and terrain effects to ar-
rive at normalized reflectance images for pho-
togrammetric multi-point positioning.
At last, we kindly acknowledge the research funds
provided by the Sinotech Foundation for Re-
search and Development of Engineering Sciences
and Technologies.
650
9. REFERENCES
Ackermann, F.,1984. High precision digital im-
age correlation. Proc. of the 39th Photogram-
metric Week, Stuttgart, pp.231-243.
Chen, L.C. and Lee, L.H., 1990. A systematic
approach to digital mapping using SPOT satel-
lite imagery. Journal of the Chinese Institute of
Civil and Hydraulic Engineering 2(1), pp.53-62.
CNES and SPOT Image, 1988. SPOT users’
handbook. Volumes 1 and 2.
Ebner, H. and Heipke, C., 1988. Integration of
digital image matching and object surface re-
construction. Int. Arch. Photogramm. Remote
Sensing 27(B11), pp.I11-534 to III-545.
Gruen, A.W. and Baltsavias, E.P., 1987. High-
precision image matching for digital terrain model
generation. Photogrammetria 42, pp.97-112.
Hannah, M.J., 1989. A system for digital stereo
image matching. Photogrammetric Engineering
and Remote Sensing 55(12), pp.1765-1770.
Heipke, C., 1992. A global approach for least-
squares image matching and surface reconstruc-
tion in object space. Photogrammetric Engi-
neering and Remote sensing 58(3), pp.317-323.
Rosenholm, D., 1987. Multi-point matching us-
ing the least-squares technique for evaluation
of three-dimensional models. Photogrammetric
Engineering and Remote Sensing 53(6), pp.621-
626.
Westin, T., 1990. Precision rectification of SPOT
imagery. Photogrammetric Engineering and Re-
mote Sensing 56(2), pp.247-253.
Wrobel, B.P.,1991. Least-squares methods for
surface reconstruction from images. ISPRS J.
Photogramm. Remote Sensing 46, pp.67-84.
Wu, J. and Chang, J.H., 1990. An algorithm
for point positioning on digital images and in
object space. Int. Arch. Photogramm. Remote
Sensing 28(5/2), pp.1196-1202.
SBS~FO0OFFONnDOOCCIH Ss