The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Voi. XXXVII. Part Bl. Beijing 2008 
growth algorithm based on least square adjustment; Researcher 
You Hongjian of Chinese Academy of Science presented a 
method of automatically extracting the outside shape of 
buildings from sparse DSM point cloud. The difficulty in 
extracting buildings is the extraction of the shape of roofs. For 
the regular roofs (herringbone, flat top, oblique top), the 
algorithms mentioned above are very effective, However for the 
complex ones, manual intervention or further processing are 
needed. 
At present, Shanghai Institute of Surveying and Mapping 
applies a more mature commercial software of Finland: 
TerraSolid to realize the classification and extraction of the data, 
which runs under the platform of MicroStation. The software 
was developed according to the classification and extraction 
algorithms presented by Axelsson et al. in Sweden, which 
comprises TerraScan, TerraModeler, TerraPhoto and many 
other modules. TerraScan is applied to data classification and 
extraction; TerraModeler is used in producing and processing 
each kind of surface; TerraPhoto is used in processing the 
original image. The topography model and building model 
established through the software need further repairing and 
processing if they are complex artificial buildings (Oriental 
Pearl, Viaduct and Flyover). 
Figure 2.The three-dimensional model of Lu Jia zhui area of 
Shanghai without the texture 
Figure 3. The three-dimensional model of Lu Jia zhui area of 
Shanghai with the texture 
4. RESONSTRUCTION OF THE THREE- 
DIMENSIONAL CITY MODEL 
Using TerraSolid module and assisted by a little manual 
intervention, a topography model and building model can be 
obtained from DSM. The topography mode is expressed by the 
triangular irregular net (TIN) and the building model by the 3ds 
format. Figure2 is the three-dimensional model of Lu Jia zhui 
area of Shanghai obtained from LIDAR data without the texture. 
In order to achieve a better effect of visualization, the 
topography model needs to superimpose the ortho-image and 
the building model need to paste the real texture. Figure 3 is the 
three-dimensional model of Lu Jia zhui area of Shanghai pasted 
with the picture of the texture. 
Shanghai Institute of Surveying and Mapping possesses 
abundant two-dimensional data of Shanghai. Through 
integrating the data with the LIDAR topography data and using 
the corresponding tools developed by the author the 
reconstruction of models of other factors can be accomplished. 
After obtaining the elevation information from the topography 
data and the plane position from the two-dimensional data, the 
horizontal factor model can be constructed on the basis of the 
two-dimensional surface factor structure. The non-horizontal 
factor model can be reconstructed by using the superimposition 
operation of the two-dimensional data and the topography 
model TIN and the end model is also expressed by TIN. 
5. CONCLUSION 
The prospect of the airborne LIDAR technology’s application is 
promising, and our country lags behind the developed countries 
relatively in the applied research of the technology. Shanghai 
Institute of Surveying and Mapping acquired the data of 20 
square kilometers in the center of Shanghai with the airborne 
LIDAR technology. Through data processing, information 
extraction and reconstruction of the three-dimensional city 
model, the institute has accumulated the experience of LIDAR 
data processing and proved that the LIDAR technology is a 
convenient, rapid and highly efficient method for acquiring the 
three-dimensional data. It has extremely vital practical 
significance to guide comprehensively the applied research as 
well as the research of data processing of airborne LIDAR 
technology in our country. 
REFERENCE 
Baatz, M. and Schape, A., 2000. Multiresolution segmentation: 
an optimization approach for high quality multi-scale image 
segmentation. Angewandte Geographische Informations 
Verarbeitung, pp. 12-23. 
[l]Luo,Z.Q., Zhang,H.R,etal, 2006. Airborne LIDAR 
technology. The informationization of National territory 
resources, pp.47-52 
Wang, J.Z., Li C. M., Lin Z. J., 2004. Discussion on acquiring 
the three-dimensional city technology. The Science of Surveying 
and Mapping, pp.71-74. 
Li, B.J.,Fang, Z.X., Ren, J.,2003. The research of extracting the 
characteristic of buildings from the laser scanning data.77je 
journal of Wuhan University(the version of information 
science), pp.65-69.