Zheng Wang 
  
that allows us to generate a TIN model with only the points that belong to the model. This contribution also allows us to 
work in a local or model-based coordinate system and make the separation of planes or building walls much easier. Also, 
the output building corners are in a format readily compatible with most GIS packages. Although it is indicated that only 
elevation data may not be enough to complete the building reconstruction in many cases, the elevation data should and 
can play a major role in the building reconstruction. Other information, such as multiple returns of LIDAR data and 
photography, will help to complete the reconstruction and can provide evidence to confirm or verify certain results. 
Some observations have been made from the test results. First, although symmetry and circularity can effectively 
eliminate many non-building edges, alone they cannot completely finish the task. It seems that the best role for 
symmetry and circularity is to pre-classify edges, i.e., to use them to quickly reduce the number of candidate building 
edges. And second, some LIDAR systems, such as EarthData Technology’s LIDAR system, collect multiple returns, 
which is a very effective tool to separate trees from buildings. Such kind LIDAR systems receive multiple returns on 
trees and single returns on buildings and open ground. When multiple returns of LIDAR data are available, the edge 
detection for building recognition purposes should be applied to the single return points only, since building surfaces 
should reflect only one return. By doing so, separating buildings from standing by trees may become a very simple and 
easy task. Further investigation is needed to explore the full use of multiple return information. Based on the results 
achieved by this approach, several future tests and developments have been considered, including: 
To fine turn processing parameters to produce better fit building surfaces and corners. 
* To test the approach using buildings of more complex shapes, such as L- and U-shaped buildings, to further 
evaluate the capability of the approach. 
* To investigate how other information, such as multiple returns and photography, can be incorporated to achieve 
better results or to confirm or verify the results generated from a surface data. 
5. REFERENCES 
Ackermann, F., (1996). Airborne Laser Scanning for Elevation Models.” in Geomatics Info Magazine. Vol. 10, pp. 24-25. 
Fritsch, D. and Amen, B. (1998). “Geometric Characteristics of Digital Surfaces: A Key Towards 3-D Building 
Reconstruction." in /SPRS Proceedings of Commission III Symposium on Object Recognition and Scene Classification 
from Multispectral and Multisensor Pixels, Columbus, Ohio, USA. Vol. 32 Part 3/1, pp. 119-126. 
Haala, N., Brenner, C. and Anders, K. (1998). “3D Urban GIS from Laser Altimeter and 2D Map Data.” in ISPRS 
Proceedings of Commission III Symposium on Object Recognition and Scene Classification from Multispectral and 
Multisensor Pixels, Columbus, Ohio, USA. Vol. 32 Part 3/1, pp. 339-346. 
Haala, N. and Hahn, M. (1995). “Data Fusion for the Detection and Reconstruction of Buildings." in Grün at al. (eds), 
Automatic Extraction of Man-Made Objects from Aerial and Space Images. Birkhauser-Verlag, Basel, pp. 211-220. 
Henricsson, O., Bignone, F., Willuhn, W., Ade, F., Kubler, O., Baltsavias, E., Mason, S. and Gruen, A. (1996). “Project 
AMOBE: Strategies, Current Status and Future Work.” in Proceedings of 18th ISPRS Congress, Vienna, Austria. B3, pp. 
321-330. 
Jähne, B. (1995). Digital Image Processing-Concepts: Algorithms, and Scientific Applications. 3rd ed., Springer-Verlag, 
Berlin/Heidelberg. 
Schenk, T. (1995). “A Layered Abduction Model of Building Recognition." in Grün at al. (eds), Automatic Extraction of 
Man-Made Objects from Aerial and Space Images. Birkhauser-Verlag, Basel, pp. 117-123. 
Wang, Z. (1998). “Extracting Building Information from LIDAR Data.” in ISPRS Proceedings of Commission III 
Symposium on Object Recognition and Scene Classification from Multispectral and Multisensor Pixels, Columbus, 
Ohio, USA. Vol. 32 Part 3/1, pp. 279-284. 
Weidner, U. (1996). “An Approach to Building Extraction from Digital Surface Models.” in Proceedings of 18th ISPRS 
Congress, Vienna, Austria. B3, pp. 924-929. 
  
964 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B3. Amsterdam 2000.