anbul 2004 
he building 
en as fixed 
eights than 
t to certain 
tion to the 
ral squared 
e obtained 
We can use 
os for the 
ed. The 3D 
© “multiple 
iting those 
Then these 
he manner 
define the 
ion values 
ints. 
veloped. In 
based on a 
pe will be 
arization is 
ructed 3D 
  
  
1s a tool to 
; described 
d to create 
| as ground 
of steps to 
ints can be 
ction and 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004 
segmentation algorithm | assumes that spatially separated 
clusters of point clouds representing separate buildings, and the 
distance between these clusters will be at least greater than the 
resolution of the dataset. 
Then, the procedure to “regularize” is described. The first step 
in regularization is to select a set of points representing a 
building, and extract those points that represent its boundary. 
This is accomplished using a modified convex hull algorithm. 
A least squares based hierarchical building squaring approach 
is introduced. A series of steps that determine parametric 
equations for building edges have been suggested. In all this, an 
assumption has been made that the edges of the buildings have 
only two, mutually perpendicular directions. Since shorter line 
segments are processed after longer lines, errors from previous 
steps are minimized. In this approach, no line segment is 
chosen as fixed and all are subject to certain levels of 
adjustment in direction and position, depending in general on 
the length of the line segment. Such hierarchical strategy 
ensures our solution to be robust to the lidar data resolution and 
the possible non-building points mistakenly included in the 
previous steps. 
Our experience shows that reliable segmentation is necessary 
for a quality building squaring outcome. Buildings with more 
than two principal directions and non-rectilinear edges need 
certain modification and adaptation of the reported hierarchical 
strategy. 
The limiting factors in this process can be the resolution of the 
data. To accurately model an urban environment, the point 
density of the 3D point clouds should be as high as possible, 
and ideally the spacing should be higher than 1 meter in X and 
Y direction. 
7. REFERENCES 
Ackermann,F., 1999. Airborne Laser Scanning — Present Status 
and Future Expectations, ISPRS Journal of Photogrammetry & 
Remote sensing, Vol.54, Issues 2-3, July, pp.64-67. 
Al-Harthy, A., Bethel, J, 2002. Heuristic Filtering and 3D 
Feature Extraction, ISPRS Commission III, September 9-13, 
2002, Graz, Austria 
Axelsson, P., 1999. Processing of laser scanner data- 
algorithms and applications, ISPRS Journal of Photogrammetry 
& Remote sensing, Vol.54, Issues 2-3, July, pp138-147. 
Baltsavias, E., 1999. A Comparison Between Photogrammetry 
and Laser Scanning, ISPRS Journal of Photogrammetry & 
Remote sensing, Vol.54, Issues 2-3, July, pp.83-94. 
Jie Shan, Aparajithan Sampath, 2004, Urban DEM Generation 
from Raw Lidar Data: A Labeling Algorithm and its 
Performance, Photogrammetric Engineering and Remote 
Sensing, in print. 
Morgan, M., Habib, A., 2002. Interpolation Of Lidar Data And 
Automatic Building Extraction, ACSM-ASPRS 2002 Annual 
Conference proceedings 
Rottensteiner,F., Briese, C.,2002, A New Method For Building 
Extraction In Urban Areas From High-Resolution Lidar Data, 
ISPRS Commission III, September 9-13, 2002, Graz, Austria 
Sampath, A., and J. Shan, 2003. Building Segmentation from 
Raw Lidar Data, ASPRS Annual Conference. Anchorage, AK, 
May 3-9. 
Schickler, W., Thorpe, A., 2001. Surface estimation based on 
Lidar, CD-ROM Proceedings of ASPRS Annual Conference, 
April 23-27, St. Louis, MO. 
Sithole, G. 2001. Filtering of Laser Altimetry Data Using a 
Slope Adaptive Filter In: Hofton, Michelle A. (Editing), 
Proceedings of the ISPRS workshop on Land Surface Mapping 
and Characterization Using Laser Altimetry, Annapolis, 
Maryland, The International Archives of the Photogrammetry. 
Remote Sensing and Spatial Information Sciences, Vol. 
XXXIV part 3/W4 Commission 111, pp.203-210. 
Vosselman, G., 2000. Slope Based Filtering of Laser Altimetry 
Data, International Archives of Photogrammetry and Remote 
Sensing, Vol. XXXIII PartB3, Amsterdam 2000, pp.935-942. 
Vosselman, G., Mass, H., 2001. Adjustment and Filtering of 
Raw Laser Altimetry Data, Proceedings OEEPE Workshop on 
Airborne Laserscanning and Interferometric SAR for Detailed 
Digital Elevation Models, March 1-3, Stockholm