Peer Reviewed Articles
29 Digitizing Satellite Imagery: Quality and Cost Considerations
Jon Leachtenauer, Kenneth Daniel, and Thomas Vogl
Abstract
Download
Full Article
The recent declassification of major U.S. satellite reconnaissance
programs offers a significant source of imagery to the
civil community. With nearly two billion square kilometers of
coverage collected over a 12-year period, a rich database of
imagery will become available to environmental researchers,
archaeologists, historians. and other users of archived imagery.
Imagery collected by the CORONA, ARGON, and LANYARD systems pre-dates Landsat and Earth Resources Technology
(ERTS) coverage and, thus, extends the historical
archive of satellite imagery by 12 years.
Unlike Landsat and ERTS imagery, however, the CORONA/
LANYARD/ARGON imagery was collected with film-return systems.
For many potential applications, it will be desirable to
place the data in digital format. This will require digitizing
the film records.
The National Exploitation Laboratory recently completed
a study designed to determine the impact of digitizing resolution
on the information content of the resultant digitized
products. A sample of imagery (duplicate positives) was digitized
with a sample of digitizers at various digitizing spot
sizes. The digitized data were displayed in softcopy, and imagery
analysts compared the softcopy images to the original
hardcopy products. Information loss was measured in terms
of the National Imagery Interpretability Scale (NWS). Results
of the study provide the basis for selection of digitizer resolution
as a function of information/bandwidth trade offs. A
brief assessment of relative costs as a function of digitizer
resolution was also made.
35 Height Determination of Extended Objects Using Shadows in SPOT Images
V.K. Shettigara and G.M. Sumerling
Abstract
Download
Full Article
Building heights were estimated with relatively high accuracy
using shadows in a set of single-look SPOT panchromatic and
multispectral images taken from the same satellite simultaneously.
Shadows cast by rows of trees in SPOT images were
first used to estimate mean heights of trees. Calibration lines
were then constructed to relate the actual mean heights of
rows of trees to the estimated heights. Using these calibration
lines, heights of some industrial buildings in the image were
estimated using their shadows with sub-pixel accuracy. The
accuracy achieved is better than three metres, or one-third
the pixel size of the SPOT panchromatic image. One of the
important challenges involved in the process was to determine
an appropriate threshold for delineating shadow zones
in the images. A technique is provided for this problem.
The technique is useful for estimating heights of extended
objects situated in flat terrains. The type of resampiing
used for overlaying a multispectral image over a
panchromatic image changes the accuracy of height estimation.
However, the change is tolerable if the heights to be estimated
are within the ground-truth data range used for
deriving calibration lines.
45 Regional Characterization of Land Cover Using Multiple Sources of
Data
J.E. Vogelmann, T. Sohl, and S.M. Howard
Abstract
Download
Full Article
Many organizations require accurate intermediate-scale land cover
information for many applications, including modeling
nutrient and pesticide runoff, understanding spatial patterns
of biodiversity, land-use planning. and policy development.
While many techniques have been successfully used to classify
land cover in relatively small regions, there are substantial
obstacles in applying these methods to large, multiscene
regions. The purpose of this study was to generate and evaluate
a large region land-cover classification product using a
multiple-layer land-characteristics database approach. To
derive land-cover information, mosaicked Landsat thematic
mapper (TM) scenes were analyzed in conjunction with digital
elevation data (and derived slope, aspect, and shaded relief),
population census information, Defense Meteorological
Satellite Program city lights data, prior land-use and landcover
data, digital line graph data, and National Wetlands
Inventory data. Both leaf-on and leaf-off TM data sets were
analyzed. The study area was U.S. Federal Region III, which
includes the states of Pennsylvania, Virginia, Maryland, Delaware
and West Virginia.
The general procedure involved (1) generating mosaics of multiple scenes of leaves-on TM data using histogram equalization methods; (2) clustering mosaics into 100 spectral classes using unsupervised classification; (3) interpreting and labeling spectral classes into approximately 15 landcover categories (analogous to Anderson Level 1 and 2 classes) using aerial photographs; (4) developing decision-making rules and models using from one to several ancillary data layers to resolve confusion in spectral classes that represented two or more targeted land-cover categories; and (5) incorporating data from other sources (for example, leaf-off TM data and Naeytional Wetlands Inventory data) to yield a final land-cover product. Although standard accuracy assessments were not done, a series of consistency checks using available sources of land-cover information were conducted to evaluate the effectiveness of this approach for generating accurate land-cover information for large regions.
59 The Accuracy of Vegetation Stand Boundaries Derived from Image Segmentation
in a Desert Environment
Andrès M. Abeyta and Janet Franklin
Abstract
Download
Full Article
Line intercept sampling was used to determine if boundaries
between desert scrub vegetation stands corresponded with
boundaries between regions in an image derived from a segmentation
algorithm applied to Thematic Mapper (TM) data
for the Anza-Borrego Desert State Park, California. An image
segmentation algorithm developed by Woodcock and Harward
(1992) was applied to images comprising TM bands 3,
4, and 5, from April 1987, principal components images
based on April 1987 and June 1990 imagery, and each with
texture added. The Global Positioning System (GPS) was used
to determine coordinates of both physiographic (land-form)
and vegetation boundaries in the field as they intersected
line transects. These boundary locations were then registered
to the segmented images.Image region boundaries that fell
within e tolerances (spatial error bounds) of surveyed boundaries
were considered accurate. Image region boundaries
showed less than 10 percent omission error but about 50 percent
commission error when compared with the true locations
of vegetation and physiographic boundaries. The use of
image principal components and texture in the segmentations
did not produce anticipated increases in the correspondence
between field-mapped and image-region vegetation
boundaries, although there is some suggestion that multidate
principal components may be sensitive to vegetation boundaries,
and texture to physiographic boundaries.
| Top | Home |