. Istanbul 2004
solution)!
lution) |
y
/
/
Zz
/
/
E UIDI
120 140
e classification
| of the process.
lassification of
nce. These two
he presentation
ip.
make the curve
we must keep
ion of the real
ore, the result-
attractors may
after just need
» the final bald
ion of the alti-
ight (Amiens),
zhboring aver-
route depends
check the dis-
ferent starting
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004
Figure 9: Aspect of the propagation mechanism over a valley in
Montmirail (left picture) and Amiens (right picture). The pro-
cessing time is coded in a gray level scale: first to last point pro-
cessed (black to white)
7 CONCLUSION AND FUTURE WORK
We have presented in this paper an efficient algorithm for classi-
fying a cloud of 3D points. This classification divides laser points
into three main classes: ground, low non ground and non ground
points. When performing the classification, an initial estimation
of the ground is calculated. This surface is then the input of a
deformable model algorithm. The final DTM may have a very
high resolution, and bring to the light relevant geomorphological
features such is of special interest for hydrological applications.
In a near future, we will focus our researches toward two partic-
ular points:
i. finding automatically the optimal neighborhood size de-
pending on the point density.
ii. going further into the classification process, that is study-
ing more precisely the non-ground class. With laser mul-
tiple echoes, and a local statistical approach, vegetation
can be separated from buildings. We may think as well of
using intensity measurements for clustering points within
the classified point cloud as well as optical images.
We would like at last to test this algorithm over photogrammetric
derived Digital Elevation Model, since entirely automatic meth-
ods do not work properly for extracting non-ground points.
8 ACKNOWLEDGMENTS
The authors would like to thank the Maison de la télédétection,
affiliated to the French institute for water resources (CEMA-
GREF) in Montpellier, France, for providing laser data over Rou-
jan as well as scientists of the MATIS laboratory for all fruitful
discussions we had together.
REFERENCES
Axelsson, P., 1999. Processing of laser scanner data - algorithms
ans applications. ISPRS Journal of Photogrammetry and Remote
Sensing 54, pp. 138-147.
Bretar, F., Pierrot-Desseilligny, M. and Roux, M., 2003. Estimat-
ing intrinsic accuracy of airborne laser data with local 3D-offsets.
Vol. XXXIV, International Archives of Photogrammetry and Re-
mote Sensing, Dresden, Germany, pp. 20-26.
Eckstein, W. and Munkelt, O., 1995. Extracting objects from
digital terrain models. Vol. 2572, Proc. Int. Society for Opti-
cal Engineering: Remote Sensing and Reconstruction for Three-
Dimensional Objects and Scenes.
Elmqvist, M., 2002. Ground surface estimation from airborne
laser scanner data using active shape models. Vol. XXXIV part
3A, International Archives of Photogrammetry and Remote Sens-
ing, pp. 114-118.
Elmqvist, M., Jungert, E., Lantz, F., Person, A. and Soderman,
U., 2001. Terrain modelling and analysis using laser scanner data.
Vol. XXXIV part 3/W4, International Archives of Photogramme-
try and Remote Sensing, pp. 219-226.
Fua, P., 1997. RADIUS: Image Understanding for Intelligence
Imagery. Morgan Kaufmann, chapter Model-Based Optimiza-
tion: An Approach to Fast, Accurate, and Consistent Site Model-
ing from Imagery.
Fua, P. and Leclerc, Y. G., 1994. Using 3-dimensional meshes
to combine image-based and geometry-based constraints. In:
ECCV (2), pp. 281-291.
Kilian, J., Haala, N. and Englich, M., 1996. Capture and eval-
uation of airborne laser scanner data. Vol. XXXI, International
Archives of Photogrammetry and Remote Sensing, pp. 383-388,
Part B3.
Kraus, K. and Pfeifer, N., 1998. Determination of terrain models
in wooded areas with airborne laser scanner data. ISPRS Journal
of Photogrammetry and Remote Sensing 53, pp. 193-203.
Lee, H. S. and Younan, N. H., 2003. DTM extraction of lidar
returns via adaptative processing. IEEE Transactions on Geo-
science and Remote Sensing 41(9), pp. 2063-2069.
Li, S., 1995. Markov random field modeling in computer vision.
Springer Verlag.
Lohman, P., Koch, A. and Schaeffer, M., 2000. Approaches to the
filtering of laser scanner data. Vol. 33, International Archives of
Photogrammetry and Remote Sensing, Amsterdam, pp. 540—547,
Part B3.
Metaxas, D. N. and Kakadiaris, I. A., 2002. Elastically adaptative
deformable models. IEEE Transactions on Pattern Analysis and
Machine Intelligence 24(10), pp. 1310-1321.
Montagnat, J., Delingette, H., Scapel, N. and Ayache, N., 2000.
Representation, shape, topology and evolution of deformable sur-
faces. Application to 3D medical image segmentation. Technical
Report 3954, INRIA.
Sithole, G., 2001. Filtering of laser altimetry data using a slope
adaptative filter. Vol. XXXIV-3/W3, International Archives of
Photogrammetry and Remote Sensing, Annapolis, pp. 203-210.
Sithole, G. and Vosselman, G., 2003. Comparison of filtering
algorithms. Vol. XXXIV-3/W3, International Archives of Pho-
togrammetry and Remote Sensing, Dresden, pp. 71-78.
Sohn, G. and Dowman, 1., 2002. Terrain surface reconstruction
by use of tetrahedron model with the MDL criterion. Interna-
tional Archives of Photogrammetry and Remote Sensing, Graz,
pp. 336-244.
Vosselman, G., 2000. Slope based filtering of laser altimetry data.
Vol. XXXIII, International Archives of Photogrammetry and Re-
mote Sensing, amsterdam, pp. 935—942, Part B3.
Zhang, K., Chen, S.-C., Whitman, D., Shyu, M.-L., Yan, J. and
Zhang, C., 2003. A progressive morphological filter for removing
nonground measurements from airborne lidar data. IEEE Trans-
actions on Geoscience and Remote Sensing 41(4), pp. 872-882.
Zinger, S., Nikolova, M., Roux, M. and Maitre, H., 2002. 3D
sampling for airborne laser data of urban areas. International
Archives of Photogrammetry and Remote Sensing, Graz, Aus-
tria, pp. 418-423.
