238
Data Distribution - Satellite Products
Of the vast amount of satellite imagery now
stored in archives, only a limited percentage
is ever ordered by end-users. One of the
reasons may be that the distributors provide
only limited data alternatives to the user
who generally needs a digital orthoimage and
possibly an elevation model over a specific
map section imaged in a specific timeslot.
For satellite data to be used more broadly,
distribution capabilities will need to be
vastly improved. Multi-temporal digital
datasets representing orthoimagery over
standard map sheet areas in standard map
projections will have to be offered at
affordable prices. These datasets will need
to be normalized to standard sensor,
atmospheric and illumination characteristics,
and updates will have to be available
regularly and without delays.
Until the user community can rely on a
regular data supply of consistent quality at
competitive prices it will continue to be
hesitant to set up infrastructures which
depend on this data. At present the primary
commercial data distribution channels are
EOSAT and SPOT-Image, both of which are
directly aligned with a single satellite
source (Landsat and SPOT respectively).
Similar agencies are being set up in other
countries as well. For example, in Canada,
Radarsat International (RSI) will distribute
processed Landsat and SPOT data and act as
the primary commercial distribution channel
for Radarsat data around the world.
In addition to product distribution these
agencies negotiate on behalf of their
governments the data rights granted to
individual ground reception stations. The
stations then orchestrate further
distribution in their region of interest.
Generally however, distribution by
groundstations is subsidized for the common
good rather than being based on sound
commercial principles, and is limited to the
satellites handled by the ground station.
From a user viewpoint, one would like to be
able to approach a single agency or company
and order the datasets of interest,
regardless of satellite. The proliferation
of satellites over the next few years will
hopefully encourage governments and private
industry around the globe to respond to this
need and establish better distribution
channels. In this regard it is encouraging
to see that the intent of the EOSDIS program
in the U.S. is to distribute data from a
number of satellites (be it U.S. only). As
well, companies such as Radarsat
International will act as a single-source
commercial distributor for imagery from
multiple satellites.
Data Distribution - Products
The situation for airborne sensors is
different to that of satellites in that
distribution is generally carried out by the
company flying the sensor and data is
generally collected for a specific customer.
However, it is similar in terms of the lack
of availability of standardized map products
offered by these companies. Here also the
providers of the service will have to improve
their offerings to reach the available market
potential. In Canada this has been
recognized by industry and government
participants in the MEIS-III program. The
commercialized system will output standard
data products in the form of calibrated
digital orthoimages. Providing data products
at prices the market can bear is a major
objective of this program.
Data Extraction Tools
Another factor contributing to the under
utilization of remotely sensed imagery is the
lack of integrated tools available to extract
all embedded feature, elevation and thematic
information from the imagery and to deposit
it in the application oriented data bases
with accurate geographic coordinates.
Recent years have seen a tremendous growth in
geographic information systems. Many of
these systems started as a solution to a
specific problem rather than being designed
as a general tool, and consequently each one
appears to be more suitable for certain
applications than for others. Most systems
are based on proprietary data structures,
which has complicated the ability to exchange
information between them.
At the same time, a proliferation of image
analysis systems has occurred, resulting in a
spectrum which ranges from very simple PC-
based solutions to multi-million dollar image
mapping systems. While the simpler systems
allow the imagery to be transformed in
desired map projections through simple
warping techniques, the complex ones achieve
a much higher accuracy by geocoding the
imagery using sensor position information, as
well as providing sophisticated information
extraction techniques.
So far however, remote sensing systems and
geographic information systems have developed
along separate tracks and the potential
remote sensing can offer as a reliable data
source for geographic information systems has
been largely untapped. While some suppliers
have offered links between the two types of
systems, few if any have offered integrated
solutions that work.
Realizing that effective use of remotely
sensed data has been limited by the lack of
integrated tools, Canadian Government and
Canadian Industry have collectively funded
the development of what is called a
Geographic Information Extraction System
(codename Geomate).