International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004
bridges can be a problem. This can be clearly seen in fig-
ure 9 which is an enlarged portion from the top left hand
corner of figure 6. Passing horizontally through the mid-
dle of figure 9(a) is a row of trees delineating a creek. The
road to be detected in the image runs approximately north-
south. At the intersection of the road and the creek there is
a bridge which has clearly not been detected in the binary
classification image seen in figure 9(b).
S CONCLUSION AND FUTURE WORK
5.1 Conclusion
This paper describes an effective and simple method for the
detection of roads from LIDAR data using a hierarchical
rule based system. The accuracy of the method has been
shown to be on par with the better algorithms published.
Results from this method can be expected to be equal to the
quoted accuracies or better in non-industrial or commercial
areas. The presence of many car parks and private roads
has reduced the achieved correctness value due to the high
presence of FP extractions.
a 1 , : 4
ml LL. gj dd
d i EL 5 4)
| i Pl eee. Pe
; 9 Lm
is x | PE d
dE:
Figure 8: Centrelines.
(a) Bridge crossing the creek (b) Missing road section.
Figure 9: Elevated road problem
5.2 Future Work
Improvements on the detection method are sought whilst
the automatic extraction and vectorisation of the raw LI-
DAR data into approximate road straight, curve and spiral
design primitives is a high priority.
REFERENCES
Akel, N. A., Zilberstein, O. and Doytsher, Y., 2003.
Automatic DTM extraction from dense raw LIDAR
data in urban areas. In: Proc. FIG Working Week
http://www.ddl.org/figtree/pub/ (accessed 20 Feb. 2004).
Clode, S. P., Zelniker, E. E., Kootsookos, P. J. and Clark-
son, I. V. L., 2004. A Phase Coded Disk Approach to Thick
Curvilinear Line Detection. In: EUSIPCO.
Hatger, C. and Brenner, C., 2003. Extraction of Road
Geometry Parameters from Laser Scanning and Existing
Databases. In: http://www.isprs.org/commission3/wg3/
(accessed 20 Feb. 2004).
Heipke, C., Mayer, H., Wiedemann, C. and Jamet, O.,
1997. Evaluation of Automatic Road Extraction. In:
IAPRS, Vol. 32, pp. 47—56.
Hinz, S. and Baumgartner, A., 2003. Automatic Extraction
of Urban Road Networks from Multi-View Aerial Imagery.
ISPRS Journal of Photogrammetry and Remote Sensing
58/1-2, pp. 83-98.
Huber, R. and Lang, K., 2001. Road Extraction from
High-Resolution Airborne SAR using Operator Fusion.
In: Proc. International Geoscience and Remote Sens-
ing Symposium http://www.cosy.sbg.ac.at/ reini/reini.html
(accessed 20 Feb. 2004).
Lee, H. Y., Park, W., Lee, H.-K. and Kim, T.-G., 2000.
Towards Knowledge-Based Extraction of Roads from 1m-
resolution Satellite Images. In: Proc. IEEE Southwest
Symposium on Image Analysis and Interpretation, Austin,
U.S.A, pp. 171-176.
Rieger, W., Kerschner, M., Reiter, T. and Rottensteiner,
F., 1999. Roads and Buildings from Laser Scanner Data
within a Forest Enterprise. In: IAPRS, ISPRS Workshop,
Vol. 32, La Jolla, California, pp. 185 — 191.
Rottensteiner, F., Trinder, J., Clode, S. and Kubic, K.,
2003. Building dectection using LIDAR data and multi-
spectral images. In: Proceedings of DICTA, Sydney, Aus-
tralia, pp. 673—682.
Steger, C., 1996. Extracting Curvilinear Structures: A Dif-
ferential Geometric Approach. In: Proc. Fourth European
Conference on Computer Vision, Vol. 1064, pp. 630—641.
Vosselman, G., 2002. On the Estimation of Planimet-
ric Offsets in Laser Altimetry Data. In: IAPRSIS, Vol.
XXXIV/3A, pp. 375—380.
Wiedermann, C. and Hinz, S., 1999. Automatic Extraction
and Evaluation of Road Networks from Satellite Imagery.
In: IAPRS, Vol. 32, pp. 95-100.
ACKNOWLEDGEMENTS
This research was funded by the ARC Linkage Project
LP0230563 and the ARC Discovery Project DP0344678.
The Fairfield data set was provided by AAMHatch, Queens-
land, Australia. (http://www.aamhatch.com.au)
KEY W
ABSTR
This pa]
building
Finland.
measure
allowing
made. P
against €
of build
accuracy
mainly
pulse da
18 em (1
6 cm us:
from firs
Airborne
the obje
models.
containi
points. ”
digital e.
3D city
mapping
railway
the ALS
Quality
years (e
(2004) I
analyzin
interact
measure
reported
position
(2000).
two met]
invariant
moment:
building