features are developed for classification tasks in urban areas and 
deal with the extension of objects (e.g. buildings, vegetation) in 
all three dimensions. Thus, we assume to benefit not from all of 
them for our special test data and expand the model by 
additional features. In particular we add features based on the 
average height and the curvatures concerning the classification 
task of mussel bed detection. 
From the group of features we identify a representative set for 
our classification task by a correlation-based approach out of 
the WEKA data mining software (Witten and Frank, 2005). 
Therefore, we introduce a fully labelled point cloud and use a 
consistency subset evaluator with a greedy forward search. A 
detailed description of the correlation-based feature selection 
for machine learning can be found in Hall (1999). With this 
analysis tool eight of the 26 introduced features are indicated to 
be essential for the classification task. They are described in 
more detail in the following. 
For the classification of water, mudflat, and mussel bed we use 
the eight features 
intensity 
point density 
distance to ground 
average height 
difference of average heights for various radii 
lowest eigenvalue 
Gaussian curvature 
mean curvature 
Apart from the intensity of backscattered signal, all features are 
derived from the local geometry of point distribution. Therefore, 
we use a volumetric approach and define a vertical cylinder 
with a predefined radius r to find adjacent points. The radii for 
the neighbourhood definition are set to r, 2 3m and nr, = 
10 m, depending on the features. 
The point density indicates the distribution of the LiDAR data. 
It corresponds to the number of backscatter signals per area 
(r = 3m). Especially on water surfaces, specular reflections 
(dependent on the incidence angle) can cause a significantly 
decreasing point density. For mussel bed detection the 
difference of a point and the lowest point elevation value within 
the cylinder (r = 10 m), depicted as distance to ground (dg), 
characterizes the greater elevation of this class (Fig. 3a). 
r=10m 
* xo» » wo» uw» 9 X 
      
  
  
(b) 
Figure 3: Sketch of the features distance to ground dg (a) and 
difference of average heights dh (b) in a laser point profile 
    
Further height-based features are the average height (h2) of all 
adjacent points in a neighbourhood (r = 3m) as well as the 
difference of average heights (dh) for various radii (r, = 3m, 
T? — 10 m) (Fig. 3b). 
For the determination of point's deviation from a plane, we 
calculate the three eigenvalues (A; = A, = A3) within the 
cylindrical neighbourhood based on the covariance matrix of the 
3D coordinates set up for each point and introduce the lowest 
eigenvalue À as classification feature. 
Moreover, we calculate the maximum and minimum of the 
normal curvature at a point on this plane, denoted as principal 
curvatures kq and k,. The product of the principal curvatures is 
called the Gaussian curvature K = k, * k,, the mean curvature 
H == (ky + k;) can be calculated by the mean arithmetic 
curvature. Both values, the Gaussian and the mean curvature, 
are introduced in our classification approach. 
4. EXAMPLES 
For the evaluation, our classification method was applied to a 
test data set (cf. Section 4.1). The classification results were 
compared to a reference that was generated by delineating water 
and mussel bed considering ground truth data and an 
orthoimage. The fully labelled point cloud is shown in Figure 4. 
For our presented supervised classification approach, a training 
step is necessary to learn the parameters. Thus, we divided the 
test data set into two parts and use a cross-validation for the 
classification task. Thereby, the parameters are learnt on one 
half of a test site and tested on the other one. The classification 
accuracy is assessed by the completeness and the correctness of 
the results. In order to test our CRF method, we compared the 
results to those obtained by a Maximum  Likelihood 
Classification. 
4.1 Datasets 
The test site covers parts of the German Wadden Sea in the 
southeaster part of the North Sea. It is located in the south of the 
island Spiekeroog. The test site contains a big water-filled 
tideway from west to east where no backscatters are recorded in 
some parts due to specular reflection on the water surfaces (Fig. 
4). It also includes some smaller tideways as well as mussel bed 
in the northern path. 
The data were acquired using a RIEGL LMS-Q560 LiDAR 
system during a Wadden Sea monitoring on 19.2. and 20.2.2011 
at low tide. The total area of the test site is about 0.3 km x 1.1 
km and includes 585,109 points, which means an average point 
density of about 1.6 points/m?. Information about 3D 
coordinates and intensity are available for the backscattered 
signal of each laser pulse. 
  
Fig. 4: Labelled point cloud for the dataset in the Wadden Sea 
with the three classes wafer (blue), mussel bed (red), and 
mudflat (yellow) 
   
     
42 
For t 
the c 
comp 
the-a 
to ev 
CRF 
inves 
class; 
neigl 
depic 
For 7 
more 
incor 
corre 
Class 
In co 
areas 
71%) 
leads 
Likel 
trans 
diffe; 
diffe 
the s 
traine 
How 
neigl 
incre 
inter: 
wate 
(41% 
Maxi 
For 
parti 
that « 
com] 
samp 
Mor: 
featu 
featu 
elevz 
Thes 
misc 
(479 
  
Tabl 
Corr 
and 1