user
base
the
his
ages
ility
tally
data
The
oves
the
ruly
| set
play
e in
“the
full
'hile
data
to: 2
atial
be
lace
h‘a
d (a
logy
way
shed
and
very
ling,
lerse
nust
GIS
| the
uery
mic
view
the
lace
and
w of
) the
view
point. At any time, the user may query any
point within the perspective image and
retrieve geographic coordinates, database
coordinates, the values of the color in rgb,
and the values or attributes of associated
GIS layers at that exact same location.
When multiple GIS raster layers are
available, the user can select either the
original true color image, the original GIS
layers (such as landcover), or analysis layers
such as wood duck habitat to drape over the
terrain. The user can then check the validity
of the layers by the way they correspond to
the landscape form. For example, Landsat
image data in true color may be overlaid
with Corps of Engineers 100 year flood plan
information with the result being shown in
perspective. If the flood plain does not
correspond to topographic lows near the
river, then the observer will immediately
pick out the conflict.
VI. THE OLYMPIC VILLAGE TEST
CASE
Georgia Tech assisted the Atlanta Organizing
Committee in winning the 1996 Olympic
games for the city of Atlanta. The campus
of Georgia Tech will become the Olympic
Village, and many activities are now being
planned for the campus. GTRI has
constructed a high resolution geographic data
set based on true color aerial photography, 2
foot contour elevation data from planning
maps, and spatial data sets for building,
parking lot, road, and utilities locations.
These data have been supplemented by
attribute information for all major buildings
on the campus. For each building, attributes
were stored showing the name of the
building, the use of the building, the height
of the building, and the number of floors in
the building. Roads were associated with
road names. In addition to the GIS attribute
information, the footprint of each building
was located as well as the location and size
of every tree on the campus. For major
buildings on the Tech campus, photographs
were take of as many sides of the building as
could be seen without major vegetation
obscuration. Buildings other than the major
119
buildings were given a generic brick
texture which is characteristic of many
buildings on the campus.
The original data base of the Georgia
Tech campus is being extended towards
Peachtree Street on the east and
downtown on the south. Similar
photographs and information will be
gathered on most of the buildings by
graduate city planning students. After the
individual building photographs have been
taken, and correlated with the GIS
attributes given above, the images are
scanned into digital form and processed to
remove optical and perspective distortion.
Figure 1 shows the Georgia Tech campus
in a perspective view with no objects
displayed. Figure 2 shows a wireframe
rendering of the campus with building and
tree locations shown. Figure 3 shows the
campus with phototexture buildings and
trees. The black lines on figure 3 are a
crosshair :cursor. The College of
Computing was identified by the system
as being located at the crosshair position.
Figure 4 gives the name of the building,
its primary use, the number of floors in
the building, and the building height.
Geographic coordinates of the building are
also reported. By moving a mouse cursor
within the perspective image to the side or
the top of a selected building and pressing
the left button, the name of the building
and all its attributes are displayed on a
popup information screen. This
interaction is dynamic so attributes of a
number of buildings can be queried at the
same session.
VII. CONCLUSIONS
Many of the functions that image analysis
and GIS systems perform today would not
have seemed possible when large jobs
were submitted to mainframes for
overnight computer runs. We used to
justify the overnight turnaround by saying
that it gave one time to think about the
