-INFLUENCES OF DIFFERENT MATERIALS ON THE 
MEASUREMENTS OF A TERRESTRIAL LASER SCANNER (TLS) 
T. Voegtle a , I. Schwab 3 , T. Landes' 3 
3 Institute for Photogrammetry and Remote Sensing (IPF), Univ. of Karlsruhe, Germany - thomas.voegtle@ipf.uni- 
karlsruhe.de 
b INSA Strasbourg, Graduate School of Science and Technology, France - tania.landes@insa-strasbourg.fr 
KEY WORDS: Laser Scanning (LIDAR), Accuracy Assessment, Accuracy Analysis, Close Range, Buildings 
ABSTRACT: 
During the last years terrestrial laser scanning became a standard method of data acquisition for various applications in close range 
domain, like industrial production, forest inventories, plant engineering and construction, car navigation and - one of the most 
important fields - the recording and modelling of buildings. To use laser scanning data in an adequate way, a quality assessment of 
the laser scanner is inevitable. In the literature some publications can be found concerning the data quality of terrestrial laser 
scanners. Most of these papers concentrate on the geometrical accuracy of the scanner (errors of instrument axis, range accuracy 
using target etc.). In this paper a special aspect of quality assessment will be discussed: the influence of different materials and 
object colours on the recorded measurements of a TLS. The effects on the geometric accuracy as well as on the simultaneously 
acquired intensity values are the topics of our investigations. A TRIMBLE GX scanner was used for several test series. The study of 
different effects refer to materials commonly used at building façades, i.e. grey scaled and coloured sheets, various species of wood, 
a metal plate, plasters of different particle size, light-transmissive slides and surfaces of different conditions of wetness. The tests 
concerning a grey wedge show a dependence on the brightness where the mean square error (MSE) decrease from black to white, 
and therefore, confirm previous results of other research groups. Similar results had been obtained with coloured sheets. In this 
context an important result is that the accuracy of measurements at night-time has proved to be much better than at day time. While 
different species of wood and different conditions of wetness have no significant effect on the range accuracy the study of a metal 
plate delivers MSE values considerably higher than the accuracy of the scanner, if the angle of incidence is approximately 
orthogonal. Also light-transmissive slides cause enormous MSE values. It can be concluded that high precision measurements should 
be carried out at night-time and preferable on bright surfaces without specular characteristics. 
1. INTRODUCTION 
During the last years terrestrial laser scanning became a 
standard method of data acquisition for many applications in 
close range domain. Nowadays an enormous variety of working 
fields can be observed like industrial production, plant 
engineering and construction, autonomous car navigation, forest 
inventories, medical and forensic applications or film industry, 
to name only a few. One of the most important and extensive 
fields is the recording and modelling of buildings which is used 
for instance for 3D city models (e.g. planning of radio networks 
or infrastructure, tourist guide systems, environmental planning, 
etc.), architecture, archaeology or cultural heritage. 
One main aspect of object capture and modelling based on TLS 
is data quality to estimate the suitability for a specific 
application. In the literature some publications can be found 
concerning the data quality of TLS (e.g. Lichti, Harvey, 2002; 
Boehler et al., 2004; Clark, Robson, 2004; Amiri, Gruen, 2005; 
Kersten et al., 2005; Lichti, Licht, 2006; Hanke et al., 2006). In 
these papers the focus is mostly geometric accuracy or 
reflectivity of standard targets or colour tables, while in this 
paper a special aspect of quality assessment - as an extension of 
previously published investigations — will be discussed, the 
influence of different realistic object materials and object 
colours on the measurements of a TLS. Besides the geometric 
accuracy also the effects on the recorded intensity values have 
been analysed. Due to the wide field of objects which can be 
preferably scanned by TLS the selection of materials was 
concentrated on those typically for building facades. 
2. TERRESTRIAL LASER SCANNER 
For these investigations the TRIMBLE GX laser scanner was 
used. The technical specifications of this scanner are sampled in 
Table 1. 
Measurement principle 
Pulsed laser 
Wavelength 
532 nm (green) 
Max. distance 
200 m (90% reflect.) 
Accuracy 
positioning 
range 
± 12 mm (100 m) 
± 7 mm (100 m) 
std. dev. plane 
± 1.5 mm (50 m) 
± 2.5 mm (100 m) 
Point size (footprint) 
3 mm (50 m) 
Min. point distance 
3.2 mm (100 m) 
Scan field 
360° (H) 
60° (V) 
Scan rate 
< 5000 pts./sec. 
Table 1. Technical specifications of TRIMBLE GX scanner 
1061