VOLUME 75, NUMBER 6
PHOTOGRAMMETRIC ENGINEERING & REMOTE
SENSING
JOURNAL OF THE AMERICAN SOCIETY FOR PHOTOGRAMMETRY
AND REMOTE SENSING
To monitor invasive floating vegetation in Lake Okeechobee, Florida, the
NASA DEVELOP team at Stennis Space Center used
MOD09 surface reflectance product from the Moderate
Resolution Imaging Spectroradiometer to calculate
Normalized Difference Vegetation Index (NDVI) as a
time series. The project team used the Time Series Product
Tool, also produced at Stennis, to create this time
series from May 1, 2008, to September 30, 2008. Lake
Okeechobee is bordered to the north and the south by
agriculture, which introduces large phosphorus loads
into the lake. Therefore, monitoring invasive floating
vegetation, including water lettuce, hydrilla, hyacinth,
and algal blooms, is of great importance due to the
potential threat they pose to the ecosystem. The image
of the lake on the far left shows the total sum of all NDVI
values for the time period studied. The middle image
shows the maximum NDVI values for the entire month
of May 2008; the highest values are red. The image on
the far right is from July 2, 2008; the brightest green depicts
areas with the highest NDVI values. This data can
assist water managers in establishing fl oating vegetation
mitigation plans. For more information contact the DEVELOP
program at NASA-DL-DEVELOP@mail.nasa.gov.
An approach to model transportation linear objects in a 3D
space and estimate their 3D lengths using planimetric road
centerline data and elevation data (Lidar and NED) and quantify
the accuracy.
A region-based segmentation developed, tested, and evaluated
for automatically detecting roads, buildings, and other
man-made objects from aerial and satellite images.
The prerequisites for an interpolation function to preserve
topographic orderliness, i.e., if point A is higher than point
B, the interpolated elevation of A should remain higher, are
revealed for the first time.
The 6S code provides a signifi cantly lower radiometric variation
(2.8 percent) than the use of pseudo-invariant features
(4.1 percent), which remains a valid approach with only a
few carefully selected invariant sectors.