1065 
■ . 
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008 
decrease, from MSE = ± 7.0 mm at 100° (orthogonal direction) 
to MSE about ± 1.0 mm below 20°. However, it has to be taken 
into account that in the same way as the MSE values also the 
number of points decrease and therefore, the estimations may 
be too optimistic. Another interesting aspect is the dimension of 
the MSE for orthogonal data acquisition or large incidence 
angles. These values of about ± 7.0 mm to ± 4.3 mm lie 
significantly above the general accuracy of this laser scanner 
which had been confirmed by the other tests during these 
investigations. 
6.2 Intensity analysis 
The acquired mean intensity values for every incidence angle 
are assembled in Table 3. Due to the high reflectivity of metal 
applying orthogonal scanning direction total reflection occur 
and the intensity values lie in the saturation domain. Expectedly 
the values decrease drastically with smaller incidence angles. 
Incidence 
Mean Intensities 
Number 
Angle [ gon ] 
|grey values] 
of points 
100 
253.8 
63 602 
90 
231.4 
72 086 
80 
149.7 
66 779 
70 
99.3 
65 247 
60 
68.2 
56 756 
50 
49.8 
51 783 
40 
36.8 
44 214 
30 
28.8 
33 791 
20 
16.6 
23 424 
10 
9.1 
11 009 
Table 3. Mean intensity values of a metal plate for different 
angles of incidence 
7. ANALYSIS OF PLASTER MATERIALS 
For analysing the reflection characteristics of plasters some 
typical types had been provided. Therefore, several plates of 
plasters with known particle sizes were available. Four types 
with particle sizes of <1.5 mm, 1.5 mm, 2-3 mm and 3 mm 
had been used for these investigations. Due to the same colour 
of all test plates (white) only the range measurements are of 
interest. Again the measurements have been executed by day 
and night-time. The obtained results are depicted in Figure 10. 
Least square adjustment 
2,44 
2,56 
2,70 2,69 
O Day 
■ Night 
particle <1,5mm 1,5 mmm 2-3mm 3mm 
Particle size 
Figure 10. Range accuracy for different types of plaster 
Certain differences can be recognised between the 
measurements of day and night-time where the MSE values 
captured at night-time are again somewhat smaller and show 
more significantly a dependence on the particle sizes, i.e. the 
particle sizes affect the measurements systematically which is 
predominantly caused by the varied roughness of the structured 
surfaces. The MSE values for day time increase with the 
particle sizes only from ± 2.3 mm (particles < 1.5 mm) up to ± 
2.7 mm (particles of 3 mm) but from ± 1.8 mm - which 
corresponds to the basic accuracy of the laser scanner - up to ± 
2.7 mm for night-time acquisition. Therefore, it can be 
distinguished between the plaster types by analysing the range 
measurements and their MSE values. 
8. ANALYSIS OF LIGHT-TRANSMISSIVE MATERIAL 
A special aspect of these investigations was the characteristic of 
light-transmissive materials (films) where different levels of 
transparency were used. Three slides had been centred in front 
of the laser scanner with a larger distance of about 3 m to the 
wall in the background in order to be able to exclude reflections 
from the background. The three slides had a transparency of 5%, 
20 % and 35% respectively. As in the case of plasters only the 
range measurements can be reasonably analysed. Figure 11 
illustrates the MSE values for these slides. 
Least square adjustment 
Figure 11. Range accuracy for slides of different transparencies 
Even after elimination of outliers (e.g. from the background) the 
MSE values of range measurements are worse about a factor of 
20 compared to the results for non-transparent materials, an 
effect which can be explained by the noise which may be 
probably introduced by transparent materials. Furthermore a 
strong correlation between MSE value and transparency can be 
observed. The slide with 5% causes the highest values of about 
± 34 mm which decrease to ± 15 mm at 35% transparency. 
9. ANALYSIS OF THE INFLUENCE OF WETNESS 
Since wetness as well as roughness are known to be important 
disruptive factors for active systems, their influence on laser 
beams has been investigated. The effects of wetness has been 
analysed with regard to two materials, wood and concrete wall. 
As expected, the tests confirm that wetness of wood leads to a 
significant decrease of the obtained intensity values. While the 
; V'