same manner and matching is carried out on this represen-
tation level.
Table 4: Contour related template representation
Templ. Q- Matches
type factor successful failed oo
rel. [%] | abs. | abs. | [um]
t1c/t2c 0 86.8 1488 226 2.90
t3c/t4c 0 65.4 1121 593 2.99
t5c/t6c 0 85.0 1456 258 2.92
Even though no compression is taken into account in this
test a relatively high quota of failures is obtained with all
template combinations. This can be explained by the fact
that in the gradient images noise is amplified with a nega-
tive impact on matching.
4. CONCLUSION
In this papers we presented a procedure for semi-automatic
measurement of signalized points at Digital Photogrammet-
ric Workstations. In general signalized targets may appear
in several images. Therefore, we aim at solving the target
location by adding multi-image constraint. For both, tem-
plate matching and image matching an area based least
squares solution is used. Further aspects like the varying
background problem are taken into account by the proce-
dure.
The experimental investigation with 1714 images of signals
in a block with 36 photographs have shown that the pro-
posed procedure was quite successful by measuring 99
% of all image points. Furthermore, the accuracy of this
automatic localization with 2.8 jm is 15 % better than the
296
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996
3.2 um accuracy obtained for the manual measurements.
The results obtained using compressed image data indi-
cate no significant loss of precision and success rate up to
a compression ratio of 1:10.
REFERENCES
Geiselmann, Chr., Hahn, M. [1994]: Identification and Lo-
cation of Simple Objects for Real-time Mapping. In:
International Archives of Photogrammetry and He-
mote Sensing, vol. 30, part 2, pp. 459-470.
Gülch, E. [1995]: Automatic Control Point Measurement.
In: Photogrammetric Week '95, Fritsch, D., Hobbie, D.
(Eds.), pp. 185-196, Wichmann Verlag, Heidelberg.
Hahn, M. [1993]: Measurement by Image Matching -
State-of-the-Art in in Digital Photogrammetry. In:
Photogrammetric Week '93, Fritsch, D., Hobbie, D.
(Eds.), pp. 33-42, Wichmann Verlag, Karlsruhe.
Kaiser, Chr., Wolter, K.-J., Goetzke, T. [1992]: Lageerken-
nung von PSBs. Bild und Ton, Band 45, Heft 11/12,
pp.333 -336.
Schickler, W. [1992]: Feature Matching for Outer Orien-
tation of Single Images Using 3-D Wireframe Control
Points. In: /nternational Archives of Photogrammetry
and Remote Sensing, vol. 29, B3.
Tsingas, V. [1991]: Automatische Aerotriangulation. In:
Proceedings of the 43rd Photogrammetric Week,
Schriftenreihe des Instituts für Photogrammetrie der
Universität Stuttgart, Heft 15, pp. 253-268, Stuttgart.
Vosselman, G., Haala, N. [1992] Erkennung to-
pographischer PaBpunkte durch relationale Zuord-
nung. Zeitschrift für Photogrammetrie und Fern-
erkundung, 60. Jahrgang, Heft 6, Juli 1992.
KEY \
ABSTE
Consid
sensing
comme
absolut
and the
KURZ
In den
Fernerl
für die
die aut
autom:
tischen
Image
the co
as the
compu
for ge
tation
bing s
metric
coordi
tion b
proble
try are
stage «
any te
ted he
