COMPARISON OF DIGITAL PHOTOGRAMMETRY AND LASER SCANNING 
D. D. Lichti d , S. J. Gordon 3 , M. P. Stewart 3 , J. Franke 3 and M. Tsakiri b 
a Curtin University of Technology, GPO Box U1987, Perth, WA 6845, Australia (D.Lichti@curtin.edu.au ) 
National Technical University of Athens, 9 H. Polytechniou Street, Athens 15780 Greece 
KEY WORDS: Laser scanning, calibration, benchmark testing, digital photogrammetry 
ABSTRACT: 
Terrestrial laser scanners (TLSs) are being used more frequently in cultural heritage recording due to their high data acquisition rate, 
relatively high accuracy and high spatial data density. In terms of measurement accuracy and quality assurance, performance 
evaluation techniques and tractable calibration procedures for TLSs are still evolving. This paper reports on three sets of experiments 
designed to evaluate scanner performance. The first set was conducted in a laboratory environment with the aim to quantify 
measurement precision as a function of different operational parameters such as scan resolution, pulse mode and range accuracy 
mode. Results indicate scanner precision was independent of resolution and pulse mode, but dependent upon range accuracy mode. 
In the second set, a cylindrical object was moved in known vertical increments and scanned at each location to empirically quantify 
scanner sensitivity. Displacement was estimated by comparing surfaces modelled from the scanner data. Displacements of greater 
than 8 mm were recovered with an RMS accuracy of less than ± 1mm. The third experiment was measurement of a wooden bridge 
deforming in response to an applied static load. Bridge deformations estimated by precision digital photogrammetry and laser 
scanning are compared. 
1. INTRODUCTION 
Terrestrial laser scanners (TLSs) are being used more frequently 
in cultural heritage recording due to their high data acquisition 
rate, relatively high accuracy and high spatial data density. 
Though parameters vary from instrument to instrument, 
acquisition rates vary from 1 to 6 kHz, reported ranging 
accuracy ranges from ±6-25 mm and sampling interval (and 
hence point density) is generally programmable. These features 
coupled with the direct measurement of three-dimensional co 
ordinates are indicative of the perceived advantages of TLSs 
over digital photogrammetry. 
The issues of accuracy, benchmark testing and calibration of 
TLSs have yet to receive serious attention. Evaluation of 
scanner performance by independent organisations has been 
limited in scope due to the relatively recent emergence of the 
technology and prohibitive hire and purchase prices. Extensive 
testing is required not only for quality assurance but also to 
develop appropriate and tractable field calibration procedures 
that can be performed prior to commencement of a heritage 
recording project. The authors’ experiences to date (Lichti et 
al., 2000; Gordon et al., 2001; Lichti and Harvey, 2002) have 
indicated that a TLS—a uniform sampling measurement 
device—is a fundamentally different technology from surveying 
instruments and digital cameras. As such, a TLS cannot be 
calibrated or evaluated using standard methods that rely upon 
the use of precisely measured baselines or target arrays. 
This paper presents the results of three sets of tests conducted 
using Curtin University’s I-SiTE TLS. The first tests were 
performed in a laboratory to assess TLS performance at close 
range. The aims were to quantify precision as a function of the 
several different operational modes of the scanner in a 
controlled environment. The second set of experiments was 
performed to quantify scanner sensitivity in controlled lab 
conditions. A cylindrical object was repeatedly scanned after 
being moved in known increments. The third experiment was 
comparative testing of precision digital photogrammetry and 
TLS measurement. This was conducted on wooden bridge 
undergoing a series of structural load tests. Here, the aim was 
to quantify scanner sensitivity on a real job site. 
2. LABORATORY TESTING 
2.1 The Scanner 
Curtin University took delivery of its I-SiTE TLS system in 
early 2002. At the core of the system is a Riegl LMS Z-210 
scanner that offers a 336° horizontal field of view (FOV) and an 
80° vertical FOV. Range accuracy of the near infrared 
rangefinder is quoted at ±25 mm in high accuracy mode and 
±50 mm in standard accuracy mode. Maximum range in each 
case is 350 m and 700 m, respectively. The I-SiTE system 
offers four preset sampling resolutions, namely coarse, medium, 
fine and ultra. The user is able to collect an individual scan of a 
scene or multiple scans that can be averaged to improve 
accuracy to ±6 mm (16 scans in high accuracy mode). Also 
available is the ability to record first pulse or last pulse range 
measurements. 
2.2 The Experiments 
The primary objective for the laboratory testing was to quantify 
the scanner’s performance in terms of range precision as a 
function of the numerous available scanning options. A flat, 
diffusely reflecting beige wall was repeatedly scanned from a 
range of 6.1 m. The FOVs were set such that a 1.7 x 1.2 m area 
was scanned at normal incidence (82° at the extents). The time 
and room temperature, which remained constant during testing, 
were recorded at the commencement of each scan. 
In total, twenty-one different scanner data sets, each consisting 
of 16 repeat scans, were acquired. The operational parameters 
Corresponding author 
-39-