ISPRS, Vol.34, Part 2W2, “Dynamic and Multi-Dimensional GIS”, Bangkok, May 23-25, 2001 
81 
3D SPATIAL OBJECTS MODELING AND VISUALIZATION BASED ON LASER LANGE DATA 
Jie DU 1 , Apisit EIUMNOH 1 , Xiaoyang CHEN 2 , Michiro KUSANAGI 2 
Natural Resources Management, 
Rural Development, Gender and Resources Program, 
Asian Institute of Technology, 
P.O. box 4, Klong Lung, Pathumthani, 12120, Thailand 
Space Technology Application and Research Program, 
Asian Institute of Technology, 
P.O. box 4, Klong Lung, Pathumthani, 12120, Thailand 
KEYWORDS: 3D spatial data, visualization, laser range-scanning data, DEM, DSM 
ABSTRACT 
The laser technology has been currently brought into photogrammetry and cartography fields as a tool for mapping. The main application 
of laser scanning systems concerns large-scale and precise topographic DEMs. Automatically interpret range images for extracting geo 
spatial features and reconstruction of geo-objects are the main research problems. The objective of this paper is to develop the 
algorithms and methods for modeling and visualization of 3D spatial data in large scale based by processing laser-scanning data. For 
processing airborne laser range data, a set of algorithms should be developed. Those algorithms are mainly include: TIN based range 
image interpolation, MM (Mathematical Morphology) based range image filtering, features extraction and range image segmentation, 
feature generalization and optimization, 3D objects reconstruction and modeling, CG (Computer Graphic) based visualization and 
animation of virtual environment. 
1 Introduction 
Over the past few years, the need for describing larger scale 
3D spatial data is continually increasing. These data are used 
for variety of applications such as region planing, architecture, 
archaeology, disaster prevention, microclimate investigations or 
transmitter placement in telecommunication. Many kinds of 
raster and vector based models for describing, modeling, and 
visualizing 3D spatial data in large scale have been developed. 
With the development of laser technology and sensor 
techniques, several kinds of airborne laser scanners are 
available for acquisition of high accuracy 3-D spatial data in 
real or very fast time. The main application of laser scanning 
systems concerns large-scale and precise topographic DEMs, 
especially in areas that are difficult for conventional 
photogrammetry. These are forest areas of total or scattered 
coverage. As a side product tree heights are obtained. Other 
interesting applications concern coastal areas, wetland, 
beaches, dunes etc. As in all successful developments, new 
applications emerge which had not been anticipated at the 
beginning. 
The purpose of this research is to study and develop the 
algorithms and methods for modeling 3D spatial objects based 
on airborne laser range data. For processing airborne laser 
range data, a set of algorithms should be developed. Those 
algorithms are mainly include TIN based range image 
interpolation, MM based range image filtering, features 
extraction and range image segmentation, feature 
generalization and optimization, 3D object reconstruction and 
modeling, CG based visualization and animation of virtual 
environment. In this research we using laser scanning data of 
Kyoto Station area simulate 3D reconstruct result. 
2 Methodology 
The airborne laser scanner reaches area coverage, for 
instance, by an oscillating deflection of the laser beam 
perpendicular to the flight direction. The instant angle of 
deflection has to be known precisely, as the basic geometric 
principle is maintained of providing position, direction and 
length of the vector to the ground point, for each shot. 
An essential feature of laser scanning is the potentials for 
almost complete automation, the GPS, INS and laser data 
being digitally recorded. After the GPS processing and the 
necessary system calibration the computation of the terrain 
points, the interpolation of a DEM and the block formation of 
the DEM is quite straightforward, as far as open terrain is 
concerned. 
It includes following main problems: 
• Pre-processing for inputting airborne Laser range 
data 
• Semi-automated extraction of 3D Spatial features 
• Methods for 3D-visual modeling 
• Generation of virtual reality environments 
• Editing & Checking methods 
Figurel Integration of laser range images and existing maps 
Figure2 3D object reconstruction based on laser range images