- 93 - 
THEALASERMETRY: A HYBRID APPROACH TO DOCUMENTATION OF SITES 
AND ARTEFACTS 
Claude E. Borg and Joseph A. Cannataci* 
Key words. Laser Scanning; Theodolite Total Station; Close-Range Photogrammetry; Hybrid Approach; Cultural-Heritage; Metric 
Propagation Error 
Abstract 
The use of digital technology for measuring and recording three-dimensional shapes of high morphological complexity, as in the case 
of the pre-historic temples in Malta, offers a lasting and faithful virtual record that can be used by future researchers. This paper dis 
cusses an integrated system, based on the accuracy of the theodolite, and the technologies of laser scanning and photogrammetry by 
maximizing the data collection capacity of each instrument. This hybrid system overcomes limitations encountered when using solely 
photogrammetry or laser scanning. Photogrammetry used in combination with laser scanning leads to a more cost-effective and less 
time-consuming project. Using photogrammetry in combination with the laser scanner overcomes the metric error propagation 
which is generated when using laser scanning alone. This data acquisition and processing system can be applied to a wide range of 
applications; from the documentation of historical buildings, artefacts and monuments to site and engineering surveys. 
* Claude E. Borg mailto:clmatmar@waldonet.net.mt is the Assistant Director responsible for the Documentation Division of the Malta Centre for Restoration (MCR). 
Dr. J.A. Cannataci ¡oseph.cannataci@um.edu.mt is Chairman of the MCR’s Board of Governors and Head of the Law & Information Technology Research Unit of 
the University of Malta 
1. Introduction 
The concept underlying this paper was developed as part of the 
preparatory work underlying the establishment of the 3D Imag 
ing Department (3DID) of the Documentation Division of the 
Malta Centre for Restoration (MCR) in the second half of 2001. 
The issue arose when determining the extent of precision attain 
able with the digital photogrammetry and laser scanning equip 
ment which was being evaluated by 3DID for procurement. It 
was noted that in their 2001 Case study on Donatello’s Madda- 
lena, Beraldin, Guidi et al. 1 had chosen digital photogrammetry 
over a 3D range camera as a reference measurement system 
because the former had promised a 0.2mm accuracy range over 
a 2mm accuracy range for the cloud points produced by the 
range camera scanning system and this in a statue which is ap 
proximately 180 cm high. The metric error propagation in the 
3D range camera scanning system was induced by the require 
ment to “stitch” 2 together several scans of the same object in 
order to produce a complete 3D model. Beraldin et. al resolved 
the problem by using non-impeding optical targets placed 
around the object to obtain reference points for the digital pho 
togrammetry to operate with. 
One of the primary applications of 3D imaging for MCR is that 
of surveying relatively large megalithic temple complexes and 
other architectural edifices where the necessity of stitching sev 
eral scans together was evident ab initio. Therefore the issue of 
metric error propagation was immediately taken into account 
when planning 3DID’s systems since the procurement of hard 
ware and software for both digital photogrammetry and laser 
1 J.-A. Beraldin, G.Guidi, S.Ciofi & C. Atzeni, Improvement of Met 
ric Accuracy of Digital 3D through Digital Photogrammetry. A case 
study: Donatello’s Maddalena, 
http://www.vit.iit.nrc.ca/References/NRC-44911 .pdf 
2 “Stitching” denotes the assembly, automated or manual, of several 
single scans of a site or an object to produce one single 3D model of 
a large site or artefact. 
scanning was undertaken simultaneously as part of a holistic 
approach to documentation solutions in the cultural heritage 
sector. 
It was clear that MCR’s intention to use the 3D surveying sys 
tems for complex temple sites as opposed to single statues 
posed additional problems to those inherent in the case study 
tackled by Beraldin et. al. In large-scale close-range photo 
grammetry, such as that required for a temple site, it was obvi 
ous that the photogrammetrical exercise alone would also re 
quire the “stitching” together of many images thus risking met 
ric error propagation without even considering the probems of 
the 3D range camera. On the other hand, at least one of the 
leading contenders for the MCR’s laser scanning system relied 
on targets whose geometry was accurately known and recog 
nised by the restitution software which in turn used the targets 
(in this case spheres) to stitch scans together even in an auto 
mated manner. Would these targets afford MCR the time 
saving cost-effective data collection method that it was seeking 
i.e. would it enable the laser scan data and the photogrammetri 
cal data to be gathered in one single data acquisition session 
with, preferably, only one set of targets in place? This com 
bined data acquisition was notionally an attractive process since 
it was clear that however reliable photogrammetry may be for 
metric accuracy, it could never match the detailed surface mod 
elling that was possible using the cloud of points produced by a 
laser scanner. This was especially true for the first main appli 
cation area i.e. that of Malta’s megalithic temples. These offer a 
mix of a large number of regular and irregular shaped stones as 
the main constituents of the building fabric with a stone surface 
often richly textured due to weathering and erosion. What MCR 
should be looking for therefore was a system that combined 
together the best of both worlds, the speed and detail of the laser 
scanner with the metric accuracy of digital photogrammetry. 
These functional specifications were also necessary since the 
MCR’s client wished that the survey was non-task specific i.e.of