llenge
ninous
quires
private
ate the
issues
quality
vithout
ze to a
well as
a were
t of the
5, the
t of the
sibility
r that
s from
Sample
je next
n much
Juction
project
| at the
1990,
digital
oduced
iophoto
analog
ed the
equired
lefined.
analog
created
which
, new
ed new
ition of
bilities.
PRODUCTION STRATEGY
The demand for digital orthophotos increased due to the
maturity of GIS technology, the general lack of current
base cartographic data, and the dramatic drop in computer
hardware costs. ^ This sudden demand for digital
orthophoto data could not be met with existing resources
and budget levels. Pooling of government resources and
relying on private industry were practical solutions to the
budget limitations and the limited production capacity.
SYSTEM AND DESIGN CONSTRAINTS
Although the USGS had long abandoned any in-house
photogrammetric hardware development, the agency
continued software development to perform system
integration tasks. For the digital orthophoto development,
the USGS expanded its prototype software into a digital
production system because commercial software was not
available. However, the system design was for a modest
amount of production work because the production
strategy was to accomplish the majority of the work by
contracting to firms in the private sector.
The basic inputs to the present system are (1) diapositives
generated from available photography acquired from the
National Aerial Photography Program (NAPP), (2)
camera calibration data, (3) existing photoidentified
control from previous mapping projects, (4)
aerotriangulation data in different formats, (5) existing
digital elevation model (DEM) data from the National
Digital Cartographic Database, and (6) topographic maps
(fig. 1, box A).
The National Aerial Photography Program is a
cooperative program with State and Federal agencies to
acquire color-infrared or black-and-white aerial
photography over the conterminous United States on a
cyclic basis. Photography from this program is the
primary source imagery for the digital orthophoto
program and diapositives are generated from the original
aerial film. In this scenario, the aerial photography has
already been inspected using strict standards. However,
there are many instances where existing photography is
not suitable for the requirements of the orthophoto user.
In these situations, the aerial photography is acquired by
the mapping contractor as part of the digital orthophoto
contract. In this second scenario, inspection of the newly
acquired photography is required before production of the
digital orthophoto begins. The two scenarios presents
two dataflow paths through the production process. In-
house production scanning also presented two dataflow
paths because diapositive scanning was performed on two
Scanners made by different manufacturers, which required
slightly different data handling.
Since the USGS has the responsibility for camera
calibration in the United States, the necessary camera
225
calibration information is readily available for in-house
production purposes. However, the current USGS camera
calibration process does not provide the calibration results
in digital form to the contractors or to the general public.
In general, users are required to manually enter camera
calibration data into their orthophoto rectification systems.
Where feasible, photoidentified control points from
previous mapping projects are used to reduce the cost for
additional field control. The use of these data constrains
this portion of the production system to be analog
because these data are obtained entirely from analog
techniques.
When available, digital elevation data from the National
Digital Cartographic Database were extracted and used in
the orthophoto generation. However if digital elevation
data were not available, they were generated by
photogrammetric methods or derived from digitized
hypsographic data. This represented two additional
dataflow paths.
ADDITIONAL DESIGN REQUIREMENTS
In addition to producing the standard 3.75- x 3.75-minute
digital orthophoto quadrangles (DOQ’s), mosaicking of
DOQ’s into the traditional 7.5- x 7.5-minute quadrangle
format and limited hardcopy image output was necessary
for specific needs. Data delivery requirements include
archiving in a format for easy use in various GIS systems
while providing sufficient metadata for data management
purposes. These requirements also include using a
compression algorithm that is in the public domain and
available on various computer platforms, and distributing
the data on various media.
STANDARDS DEVELOPMENT
Since digital orthophoto data are to be generated by other
agencies and private contractors, there is an obvious need
for DOQ standards (fig. 1, box B and C). Currently two
standards are being developed in the United States. The
USGS has been working with participants of the National
Digital Orthophoto Program to develop standards for the
digital orthophoto (U.S. Geological Survey, 1995).
Meanwhile, each Federal agency also participates on the
Federal Geographic Data Committee to develop digital
orthophoto standards for the National Spatial Data
Infrastructure (Federal Geographic Data Committee,
1995). Unfortunately, an accepted national standard was
not available before system development began.
Furthermore, the long review process and the iterative
changes to the draft standard complicated the system
development effort and prolonged the development
period.
Digital orthophoto production by other agencies and
contractors also adds a functional requirement that the
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B2. Vienna 1996