CIP A 2003 XIX th International Symposium, 30 September - 04 October, 2003, Antalya, Turkey 
631 
f images pair 
rogrammatic 
steps from a 
le, described 
d from the 
models are 
;asurements, 
etical model 
r 3D model, 
merged and 
ce from the 
eration in a 
long range 
g with two 
flight) and 
uired using 
,RF; a short 
ime elapsed 
> measured, 
g head can 
: (step size 
; the vertical 
given up to 
available on 
weloped for 
Some laser 
>ws driving 
The software adopted was specifically developed for the 
treatment of clouds of 3D points coming from laser scanner 
(long range) and digital images by 3DVeritas, an Italian 
company founded in 2000 as a spin-off from the European JRC. 
Actually the software development is provided by JRC where 
the software core originally comes from. 
Reconstructor Software [Sequieira et al. 1999]; [Sgrenzaroli, 
Wolfart 2002] allows producing photo-realistic 3D 
reconstruction of real large-size object using laser scanning and 
still images. 
The workflow for laser data acquisition phase and processing 
using Reconstructor can be divided in the following main steps: 
viii) Data acquisition and storage; laser range data are acquire 
together with digital images and organized in project 
structure 
ix) Pre-processing; range data are pre-processed for local 
normal computation, confidence value computation and 
noise reduction. 
x) Data registration and geo-referencing; registration process 
transforms into a single reference frame, range data 
obtained from different viewpoints. This process can be 
also provided directly by laser acquisition software using 
external targets. 
xi) Texture mapping; Reconstructor peculiarity is the 
possibility to use range information for external camera 
calibration and texture mapping over range information. 
xii) Meshing; meshing tool converts the set of raw 3D points 
into a triangulated surface (mapped with reflectance or 
texture data) 
The listed steps are described for the specific case study of this 
research in section 9. 
4. PHOTOGRAMMETRIC TOOLS 
4.1 The photogrammetric way 
The photogrammetric survey is driven by three constraints: a 
general photogrammetric survey, in order to get the camera 
orientation by bundle adjustment, a specific survey driven by a 
theoretical model, and an automatic or semi automatic 
measuring method in order to represent a specific surface of the 
measured object. The first step is made using generic and 
commercial photogrammetric software. In this project this first 
step was made using PhotoModelerV4.0. Data have been 
imported in Arpenteur software for the second step, Roma 
software was use for the third one. Arpenteur and Roma 
software are described in the following sections. 
4.2 Arpenteur 
ARPENTEUR (meaning ARchitectural PhotogrammEtry 
Network Tool for Education and Research) is a set of software 
tools developed by Pierre Drap and Pierre Grussenmeyer, MAP 
CNRS laboratory. These tools are based on the notion network 
use and rely on the IP communication techniques. Examples 
can be consulted on the Internet site 
http://arpenteur.gamsau.archi.fr. 
As a tool dedicated to archaeology and architecture, 
ARPENTEUR benefits from the expertise of two teams in the 
fields of close range photogrammetry and the representation of 
architectural information. 
The main objective is founded on the idea of a process guided 
by the information related to the field. Concerning architecture 
and archaeology, the goal is to allow experts to use their 
knowledge to produce results which ideally meet their wishes. 
[BarcelQ 2000] The results can be shown as documents, visual 
files, or as a body destined for a database. For this purpose the 
system gives to the experts a set of tools which allow them to 
formulate hypotheses related to their field of investigation, 
hypotheses that lead to easier measurement process. Between 
these, for example, the creation of a body representing the 
objects in their field of investigation. 
As a benefit of those choices, the ARPENTEUR looks like a 
tool developed for professional architects and archaeologists 
with minimal intervention from the photogrammetry expert. 
4.3 Roma: 3-D automatic Measurement Principles 
ROMA, Representation of Oriented Model for Arpenteur, is a 
tool build on the I-MAGE method (standing for Image 
processing and Measure Assisted by GEometrical primitive) 
developed in the framework of the Arpenteur Project [Drap, 
Grussenmeyer, Gaillard, 2001]. Roma allows automatic 
measurement using a set of oriented photograph and a mesh 
visible on these photographs. 
Two types of solutions are offered in software market: in simple 
software this phase is reduced to the minimum, the operator 
shows a point on a photograph using to the mouse pointer and 
the homologous point on another photograph; this solution is 
used for example in: Photomodeler, Rollei CDW, etc.... In 
more sophisticated software, based on stereoscopy, sometimes 
bounded with correlation, homologous pointis determination is 
more accurate but limited to at list two photographs and it needs 
the intervention of a photogrammetric restitution professional. 
Roma uses simplified geometrical model, i.e. a surface mesh, 
image correlation and oriented photograph to determine 3D 
points visible on photograph and included in the mesh. 
We employ four steps ( ) in this Semi-automated Primitive 
Measurement Method, considering that a mesh has been 
measured and computed from a set of 3-D points visible on at 
least two images: 
X For each triangle of the mesh we scan triangle and get 
point f]- Each point f] is projected as pi on to the 
photograph 1; 
X [] is projected as p2 onto the second image; 
X Point p2 is used as an approximate position to initiate the 
area based correlation process with pi; 
X Point p3 is the result of the correlation; pi and its 
homologous p3 are used for the computation of the 3-D co 
ordinates of ni.