w 
681 
FLOATING MODEL FOR BUILDING RECONSTRUCTION FROM TOPOGRAPHIC 
MAPS AND LIDAR DATA 
Sendo Wang 
Department of Geomatics, National Cheng Kung University, 1 University Road, Tainan City 70101, 
Taiwan, China - sendo@geomatics.ncku.edu.tw 
Commission III, ThS-7 
KEY WORDS: Building, Reconstruction, LIDAR, Photogrammetry, Virtual Reality, Digital, City Model 
ABSTRACT: 
A novel approach of Model-based Building Reconstruction (MBBR) from topographic maps and LiDAR data called Floating 
Models is proposed in this paper. Floating models are a series of pre-defined primitive models which are floating in the space. Its 
size is adjustable by shape parameters, while its location and rotation is controlled by pose parameters. A building is reconstructed 
by adjusting these model parameters so the wire-frame model adequately fits into the building’s outlines among the topographic 
maps, LiDAR data, aerial photos and DEM. This model-based reconstruction provides good constraints to the shape of the model in 
contrary to the data-based approach. In this paper, the model parameters are re-arranged into two groups: plane and height 
parameters. The plane parameters are determined by fitting the top or bottom boundary of the model to the topographic maps. The 
height parameters are decided by fitting the top surface of the model to the lidar data and interpolating the datum point’s height from 
DEM. The proposed reconstructing procedure is semi-automated. First, the operator chooses an appropriate model and 
approximately fit to the building’s outlines on the topographic map. Second, the computer computes the optimal fit between the 
model and the topographic map based on an ad hoc least-squares model fitting algorithm. Third, the computer computes the roof or 
ridge height form the lidar points within the roofs boundary. Finally, the model parameters and standard deviations are provided, 
and the wire-frame model is superimposed on all overlapped aerial photos for the operator to check the result. The operator can make 
any necessary modification by adjusting the corresponding model parameter. We select a small urban area of Taipei City for testing 
the proposed approach. The fitting result is compared to the traditionally photogrammetric result. Most of the modem buildings can 
be modeled smoothly, and fitting result achieves the photogrammetric accuracy. 
1. INTRODUCTION 
In response to the development of 3D City Spatial Information 
Systems for urban planning and management, acquisition of 3D 
data of city objects has increasingly become an important topic 
(Braun et al. 1995; Englert and Guelch 1996; Gruen 2000; Lang 
and Foerstner 1996; Vosselman and Veldhuis 1999). Conven 
tional photogrammetry concentrates on the accurate 3D coordi 
nate measurement of points. The automated measuring systems 
set up by image matching algorithms are still based on the 
point-to-point correspondence. However, higher-order features 
such as linear, planer or volumetric features contain much more 
geometric and semantic information than a single point. 
The increasing demands of object’s 3D models encourage many 
researches toward using 3D CAD models as a modeling tool to 
extracting objects from image data (Bhanu et al. 1997; Boehm 
et al. 2000; Brenner 2000; Das et al. 1997; Ermes 2000; Tseng 
and Wang 2003; van den Heuvel 2000). This trend towards in 
tegration of photogrammetry and CAD system in the algo 
rithmic aspect creates a new term: “CAD-based Photogram 
metry'". Researches show that using CAD models does increase 
the efficiency of photogrammetric modeling by two reasons: (1) 
the advanced object modeling techniques such as Constructive 
Solid Geometry (CSG), (2) the incorporation of geometric ob 
ject constraints. 
Inspired by the CAD-based photogrammetry, we proposed a 
naval measuring tool - floating model - for reconstructing 
building from both 2D and 3D data. The floating model repre 
sents a flexible entity floating in the 3D space. It can be a point, 
a line segment, a surface plane, or a volumetric model. Each 
model is associated with a set of shape parameters and a set of 
pose parameters. The pose parameters determine the datum 
point’s position and the rotation of the model. The shape pa 
rameters change the model’s outline and volume. From the tra 
ditional photogrammetric point of view, the floating model is an 
extension of the floating mark. Instead, it is not only floated in 
the object space, but also deformable to fit the outline of the ob 
ject. From the model-based building reconstruction’s point of 
view, floating mark is an exceptional case of floating model 
without any shape parameter. Three kinds of primitive models - 
box, gable-roof, and polygonal prism - are designed for build 
ing reconstruction in our case study. 
Model-based building reconstruction (Ameri 2000; Brenner 
1999; Sester and Foerstner 1989; Wang and Tseng 2004) starts 
with hypotheses of building model representing a specified tar 
get on the scene, and verifies the compatibility between the 
model and the existing data, such as topographic maps, aerial 
photos, LiDAR data, and DEM. Approaches to MBBR are 
mostly implemented in a semi-automatic manner, solving the 
model-data fitting problem based on some high-level informa 
tion given by the operator. The spatial data of a building object 
are determined, when model-data fitting is achieved optimally. 
Therefore, the key is the algorithm that is able to determine the 
pose and shape parameters of a floating model such that the 
edge lines of the wire-frame are optimally coincided with the