STATE-OF-THE-ART AND FUTURE. MEE D FOR 
DEVELOPMENT OF DIGITAL TERRAIN MODELS 
Dr. Marshall B. Faintich 
Chief, Advanced Technology Division 
Systems and Techniques Directorate 
Defense Mapping Agency Aerospace Center 
St.n Louis, Missouri 63118 U.S.A. 
Commission ITTI 
ABSTRACT 
The state-of- the- art of production of digital terrzin models 
has progressed from simple digitization of contour sheets 
with subsequent digital interpolation to analytical stereo 
compilation with full photogrammetric contrel. Furthermore, 
current developmental efforts include digital terrain models 
produced by automatic digital stereo image © rrelation, 
analysis of multispectral responses influenced by terrain 
modulation, and compilation from stereos synthetic aperture 
radar imagery... In addition, theoretical studies are 
addressing geomorphological quality and accuracy estimation 
of.digital terrain models, empirical studies are add lressing 
model differences as a function of production equipment/ 
methodology, and experimental novios are developing tech- 
niques for digital terrain model anomaly detection and re- 
moval. 
INTRODUCTION 
[he task of,.compilins a comprehensive paper on the state- 
of-the-art of digital terrain models is indeed a complex 
issue... This is due primariiy because of the ever-increasing 
demand to represent terrain in a digital format for expioi- 
tation bv analysts using modern computers for a variety of 
apPIICaCIONS, including resource exploration, land use and 
mission planning, computerized galdance systems, aircraft 
and land vehicle simulators, and automated chart production. 
Because of the wide variation of source data, including var- 
ious types.of sensor data, different scales and resolution 
of input soie and magnitude of required terrain coverage, 
coupled with the variation of photogrammetric exploitation 
equipment and es resources, digital terrain model 
producers have developed a wide variety of state-of-the- 
art techniques to optimize production based on individual 
production requirements, source information, and exploita- 
tion equipment combinations 
To further enhance the complexity of the issue, similar in- 
dividual scenarios have resulted in variations In digital 
terrain model data structure, storage, and representation. 
The one common thread in the various production and repre- 
sentation methodologies for digital terrain models is the 
reauirement to mathematicallv measure the accuracy and 
quality of the models, and to develop techniques to remove