In: Wagner W„ Szdkely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B 
Figure 2: Operational workflow of radiometric calibration of full- 
waveform ALS data. 
observation stations close to the campaign site afterwards. With 
the help of radiative transfer models an atmospheric attenuation 
coefficient can be derived (see figure 2(b)). 
Decomposing the full-waveform data yields a 3D point cloud 
with additional information per echo such as range, amplitude 
and echo width (see figure 2(c)) (Wagner et al., 2006). 
From here on the radiometric calibration is further processed by 
OPALS modules that were developed by the Institute of Pho- 
togrammetry and Remote Sensing (IPF) of the Vienna University 
of Technology (IPF, 20J0). The opalslmport module is used to 
load the 3D point data with its attributes and its corresponding tra 
jectory strip-wise into the OPALS data manager system for subse 
quent use in all OPALS modules dealing with point clouds. Dur 
ing the import process, opalslmport reconstructs the beam vec 
tor, echo number and number of returns of each echo and stores 
them as additional attributes in the data manager. Furthermore, 
the opalsNormals module performs a local plane fit for each point 
based on its neighbouring points in order to derive the local nor 
mal vector of each point. However, due to the partly high surface 
variation, it might not be possible to fit a plane for every echo 
(only planes with a maximal user specified tolerance value for the 
adjusted sigma value are accepted), e.g. in case of echoes orig 
inating from vegetation. Nevertheless, if the plane is accepted, 
the normalized normal vector is stored as additional attribute for 
each point in the data manager (see figure 2(d)). 
In order to allow the radiometric calibration of ALS data the 
OPALS module opalsRadioCal was developed to firstly derive a 
mean calibration constant (see figure 2(e)). Within this step, for 
every echo within a given reference surface with given reflectivity 
and atmospheric attenuation coefficient the calibration constant is 
estimated according to the second formula in equation 1, the first 
formula in equation 2 and the formula displayed in equation 3. 
For points within the reference surface the local incidence angle 
is computed from the local normal vector and the beam vector. 
These calibration constants for each echo within a reference sur 
face are used to determine a mean calibration constant for the 
whole ALS campaign. The opalsRadioCal module applies this 
mean calibration constant to secondly calculate the calibrated ra 
diometric values for each echo. This process includes the estima 
tion of the backscatter cross section, backscattering coefficients 
and incidence angle corrected values as mentioned in section 2.1 
(see figure 2(f)). 
3 RESULTS AND DISCUSSION 
The radiometric calibration procedure was tested on a data sub 
set of the Vienna wide ALS campaign carried out at the end of 
2006 and beginning of 2007, namely on the parts of thirteen flight 
strips covering the area of the Schonbrunn palace, garden, zoo 
and surrounding living area. This particular full-waveform data 
set was acquired on the December 27 th 2006, by the company 
Diamond Airborne Sensing GmbH with a RIEGL LMS-Q560, 
which operates at a wavelength of 1550 nm. The scan frequency 
was 200 kHz, the aircraft speed above ground 150 km/h, the fly 
ing height above ground 500 m and the scan angle ± 30°. These 
settings resulted in a swath overlap of about 60 %, a mean point 
density of more than 20 measurements per square meter and a 
laser footprint size on ground of about 25 cm. The meteorolog 
ical data for modelling the atmosphere was received from three 
weather observation stations located within the city of Vienna. 
Two smaller asphalt regions, one gravel region, one building roof 
and the big asphalt regions of the parade yard of the Maria There 
sia casern in the south of the Schonbrunn gardens (see figure 3) 
were chosen as reference surfaces. Reflectances at zero angle of 
incidence between 15 % for one of the smaller asphalt regions 
up to 44 % for the gravel region were determined by the RIEGL 
reflectometer. For the parade yard in the centre of the three big 
buildings in figure 3 a reflectance of 23.5 % was measured. 
Figure 3: RGB-Orthophoto of the Maria Theresia casern in the 
south of the Schonbrunn gardens (MA41, 2010). 
The parade yard of the Maria Theresia casern in the south of 
the Schonbrunn gardens is by far the largest homogeneous area 
within the test site. Therefore, it was also used as reference sur 
face during the calibration procedure. Additionally, this area en 
ables to study the different radiometric calibration values, which 
can be seen in figure 4. The left diagram of figure 4(a) shows the 
selected echoes for the analysis of two overlapping flight strips, 
the echoes of the western strip (> 65 000) in green and echoes 
of the eastern one (> 113 000) in blue. In the eastern strip the 
parade yard is located close to the centre of the swath, while for 
the western strip it is located at the swath border. This can also 
be seen in the right diagram of figure 4(a), which shows range 
versus angle of incidence. The eastern echoes were acquired at 
angles of incidence up to 22° and the echoes of the western strip 
between 18° and 30°. With increasing incidence angles also the 
ranges increase, approximately up to 70 m. Hence, the effects 
which can be seen in the diagrams below combine the range and 
the angle of incidence dependencies. Figure 4(b) shows the orig 
inal amplitude values versus range and versus angle of incidence. 
In both cases the decrease with increasing range and angle of in-