GENERATION OF HIGH FIDELITY DIGITAL TERRAIN MODELS FROM CONTOURS 
G. Aumann and H. Ebner 
Chair for Photogrammetry and Remote Sensing 
Technical University Munich 
Arcisstr. 21, D-8000 Munich 2, Germany 
Tel: +49-89-2105 2671; Fax: + 49-89-280 95 73; Telex: 522854 tumue d 
E-mail: gabi@photo.verm.tu-muenchen.de 
Commission IV 
ABSTRACT 
Generating a digital terrain model (DTM) from a given 
set of contours has particular importance due to the wide 
availability of contour maps. To generate a high fidelity 
DTM at first a sufficient data preparation of the re- 
source data is necessary. In this paper the required steps 
for the data preparation, especially the automatical deri- 
vation of skeleton lines from the given contour set, are 
shortly described. Three approaches for generating a 
high fidelity DTM using the given contours and the 
derived data are presented. The high fidelity DTM can 
be based on a triangulated irregular network (TIN) or a 
regular grid. Practical examples are presented. The re- 
sults demonstrate the efficiency of the proposed meth- 
ods. 
Key words: digital terrain model, contours, geomorpho- 
logical information, TIN-structure, grid-structure. 
1. INTRODUCTION 
The digital terrain model (DTM) has been subject of 
research and developement for more than three de- 
cades. In the last years there is a tendency to generate 
countrywide high fidelity DTMs. In this context contour 
maps, which are available in many cases with good accu- 
racy, attach importance for DTM generation (Leberl et 
al, 1984; Clarke et al, 1982). To generate a high fidelity 
DTM at first a sufficient data preparation is necessary. 
In this paper the required steps for the data preparation, 
especially the automatical derivation of skeleton lines 
from the given contour set, are shortly described. Three 
approaches for generating a high fidelity DTM using the 
980 
given contours and the derived data are presented. The 
efficiency of the proposed methods is demonstrated 
using practical examples. 
2. DATA PREPARATION 
The recording of the contour data set is to be done either 
by photogrammetric stereo measurement or by digitiza- 
tion of existing contour maps. The digitalization pro- 
ceeds either manually, semiautomatically or automat- 
ically (Lichtner, 1987; Giebels/Weber, 1982; Yang, 
1990). After the data acquisition the data preparation is 
necessary to generate high fidelity DTMs. First of all the 
gross errors have to be eliminated (Aumann et al, 1992). 
Afterwards the data set has often to be thinned out by 
special algorithms without loosing essential information 
to obtain a proper data density for the further DTM 
generation (Bássmann /Besslich, 1989). 
For generating DTMs from a given contour set, geomor- 
phological information in form of skeleton lines, i.e. 
ridge and drainage lines, is of essential importance. 
Existing contour maps contain these lines implicitly. To 
extract this information automatically from the given 
contour set two approaches have been developed at 
Technical University Munich. 
The first method is based on raster data processing 
techniques. Treating the given contour set as a graph, 
medial axes are computed by means of suitable algo- 
rithms. Parts of the medial axes, i.e. those between two 
parts of the same contour line, are picked out, connected 
and used as the approximate skeleton lines (Aumann et 
al, 1991; Tang, 1991). 
The second 
will be sur 
detailed pre 
The autom: 
contour set 
each conto 
rection of t 
vector of tl 
jacent poly, 
tors in the c 
tion is nec 
network (T. 
all contour 
triangulatic 
gulation (R 
informatior 
The compu 
where they 
the TIN in 
these areas 
minimum s 
slope lines. 
derived ske 
  
  
Fig. 1: Con 
The gre 
A lot of te: 
shown, tha 
areas is pc