t image
ight image
EXTRACTING HIGH RESOLUTION DIGITAL ELEVATION MODELS AND
FEATURES IN A SOFTCOPY ENVIRONMENT
Charles K. Toth and Andre Templer
Center for Mapping
Department of Geodetic Science and Surveying
The Ohio State University
1216 Kinnear Road, Columbus, OH 43212-1154
e-mail: toth(a)cfm.ohio-state.edu
USA
Commission III
KEY WORDS: DEM/DTM, Feature Extraction, Softcopy Experiences, Urban Imagery
ABSTRACT
Digital elevation models play an essential role in many mapping and engineering applications. With the
introduction of softcopy systems, new technologies are available for both manual and automated DTM
extraction and feature collection. This paper reports our experiences with softcopy systems and compares
their performance and results to the traditional analytical plotter-based methods. The objective of our project
was to test the feasibility of new softcopy DTM and feature extraction techniques for high density urban and
suburban areas.
Traditionally, DTMs and features are extracted from "hard copy" aerial images on analytical stereoplotters.
This process is supported by specialized photogrammetric software for image orientation, data extraction,
and input to GIS-type database. Our approach used a softcopy photogrammetric workstation and digital aerial
images. The images were obtained by scanning aerial diapositives at a nominal resolution of 25 micrometers
on a Leica scanner. All the tests, including orientation and data extraction, were carried out on a
Leica/Helava DPW 770 softcopy workstation.'
1. INTRODUCTION
New, emerging technologies often require a large
volume of spatial data in an unconventional format.
A typical example is the construction of cellular
phone networks. In order to determine the optimal
distribution of cells, i.e., where to build the radio
transmitter stations, a careful spatial analysis is
necessary to assess and model the electromagnetic
signal propagation in the area. This process
frequently requires an unusually dense
representation of the surface including all natural
and man-made objects. Furthermore, a key element
of this data is the surface normal since signal
reflections play a very important role in the analysis
of the electromagnetic field. This necessitates a sub-
meter grid representation of the area including
terrain points, buildings, transportation structures,
tree canopies, etc. This "draped" surface data is the
primary input to the electromagnetic signal
propagation analysis program which is built around
a ray tracing algorithm. Different antenna positions
are analyzed in order to arrive at the optimal antenna
location. In an additional step, traditional GIS data
about the usage of the cellular phones, such as the
population demographics, traffic, and behavioral
patterns are considered to analyze the capacity
conditions and finalize the antenna positions. The
currency of the data and quick turnaround times are
also important aspects.
In cooperation with a cellular phone company, a
pilot project was carried out at the Center for
Mapping and the Department of Geodetic Science
and Surveying in 1995. The objective was to deliver
Im grid data covering the combined surface of
natural and man-made objects over two very dense
urban areas with sizes of approximately 2km by
! The Leica/Helava DPW 770 softcopy system was donated to the Department of Geodetic Science and Surveying, The Ohio State
University, by Leica Inc. in 1995.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996