Peer-Reviewed Article Abstracts
41 Radargrammetric Parameter Evaluation of an Airborne SAR
Image
Joz Wu and De-Chen Lin
Abstract
Download
Full Article
Digital synthetic aperture radar (SAR) images are composed of lines
oriented along-track and pixels in a cross-track direction. Because
of the motion of a SAR platform, the flight position and velocity
parameters have time-dependent characteristics. Our objective consists
in devising a model for the position/velocity parameters, then incorporating
the model into radar-grammetric data processing. The interpolative
parameter model is based on a linear prediction method with filtering.For
a monoscopic airborne SAR image having a 12.0 km swath width and
a 9.0 km along-track distance at an altitude of about 7.1 km, the
positional root-mean-square errors are calculated using the ground
coordinates available at 15 independent check points. Taking into
account any model inadequacy, the errors inherent in the SAR image
forming, and the image point measurement errors, the position accuracy
values amount to ±1.5, ±1.5 pixels horizontally, and ±1.2
pixels
vertically.
49 Fusion of High
Spatial and Spectral Resolution Images: The
ARSIS Concept and Its Implementation
Thierry Ranchin and Lucien Wald
Abstract
Download
Full Article
In various applications of remote sensing, when high spatial resolution
is required in addition with classification results, sensor fusion
is a solution. From a set of images with different spatial and spectral
resolutions, the aim is to synthesize images with the highest spatial
resolution available in the set and with an appropriate spectral
content. Several sensor fusion methods exist; most of them improve
the spatial resolution but provide poor quality of the spectral content
of the resulting
image. Based on a multiresolution modeling of the information, the ARSIS concept
(from its French name ``Amèlioration de la Rèsolution Spatiale
par Injection de Structures'') was designed with the aim of improving the spatial
resolution together with a high quality in the spectral content of the synthesized
images. The general case for the application of this concept is described.
A quantitative comparison of all presented methods is achieved for a SPOT image.
Another example of the fusion of SPOT XS (20-m) and KVR-1000 (2-m) images is
given. Practical information for the implementation of the wavelet transform,
the multiresolution analysis, and the ARSIS concept by practitioners is given
with particular relevance to SPOT and Landsat imagery.
65 Airborne Laser Mapping of Assateague National Seashore Beach
W.B. Krabill, C.W. Wright, R.N. Swift, E.B. Frederick, S.S. Manizade,
J.K. Yungel, C.F. Martin, J.G. Sonntag, Mark Duffy, William Hulslander,
and John C. Brock
Abstract
Download
Full Article
Results are presented from topographic surveys of the Assateague Island
National Seashore using an airborne scanning laser altimeter and
kinematic Global Positioning System (GPS) technology. The instrument
used was the Airborne Topographic Mapper (ATM), developed by the
NASA Arctic Ice Mapping (AIM) group from the Goddard Space Flight
Center's Wallops Flight Facility. In November, 1995, and again in
May, 1996, these topographic surveys were flown as a
functionality check prior to conducting missions to measure the elevation of
extensive sections of the Greenland Ice Sheet as part of NASA's Global Climate
Change program. Differences between overlapping portions of both surveys are
compared for quality control. An independent assessment of the accuracy of
the ATM survey is provided by comparison to surface surveys which were
conducted using standard techniques. The goal of these projects is to make
these measurements to an accuracy of ± 10 cm. Differences between the
fall 1995 and 1996 surveys provides an assessment of net changes in the beach
morphology over an annual cycle.
73 Multispectral Imagery Advanced Band Sharpening Study
Jim Vrabel
Abstract
Download
Full Article
The goal of this project was to evaluate MSI band sharpening in four
research areas. The first area explored the ``Effective Ground Sample
Distance (GSD)'' and relative utility of multi-spectral imagery sharpened
with panchromatic (pan) imagery. The second area examined interactions
between data compression and the band sharpening process. The third
area determined the effectiveness of band sharpening using a pair
of high resolution sharpening bands covering different regions of
the electromagnetic spectrum. The fourth area determined the effect
of band sharpening on the accuracy of automated exploitation algorithms
such as terrain categorization (TERCAT) and normalized difference
vegetation index (NDVI). The results of this research show that the
Effective GSD of the sharpened image can be predicted prior to fusion
(an Effective GSD equation based on the GSDs of the multispectral
and panchromatic input images is determined). The Effective GSD of
the sharpened product is shown to be close to the GSD of the panchromatic
sharpening band (94 percent of the possible GSD improvement is achieved).
The research also showed that sharpened products demonstrate increased
utility compared to the original low resolution MSI and the degree
of increased utility for sharpened products can be predicted prior
to fusion. The optimal sharpening algorithms based on this and prior
research are also discussed.
81 Requirement of Ground Tie Points for InSAR DEM Generation
Guoqing Sun, K. Jon Ranson, Jack Bufton, and Michael Roth
Abstract
Download
Full Article
Ground tie points have been used to improve the accuracy of a DEM derived
from interferometric SAR (InSAR) data. The number of tie points required
for ensuring certain accuracy of the InSAR-derived DEM was investigated
in this study in two ways. One is the least-squares estimate of the
corrections to a DEM using tie points, and the other is using an optimization
procedure to estimate the uncertainties in estimations of InSAR parameters,
mainly the baseline length, tilt angle, and reference phase using tie
points, and resultant DEM error. It was found that, in terms of required
tie points, both give comparable results. The latter method, though,
can be applied in more realistic cases to design the tie point acquisition
strategy for a mission.
87 Generating Viewsheds without Using Sightlines
Jianjun Wang, Gary J. Robinson, and Kevin White
Abstract
Download
Full Article
A new simple algorithm for computing viewsheds using gridded digital
elevation models (DEMs) is presented. By using reference planes rather
than sightlines, considerable savings in computing time can be made.
The time for computing a viewshed for every point of a DEM is constant,
in contrast to with sightline-based algorithms in which the computing
time varies with the location of a point in a DEM. The algorithm
was tested by use of the DEMs of two areas. The viewshed of every
point of both DEMs was calculated. The viewshed is, as might be expected,
found to be strongly influenced by both relative and absolute position.
Viewpoints at peaks and ridges have more visible area than the viewpoints
at pits or bottom of valleys. However, the viewshed of the viewpoints
at peaks or ridges with flat local ground surface, especially in
plateau-like areas, could be less than the viewpoints in pits or
bottom of valleys.
91 Matching Road Edges in Stereo-Image Sequences Using Data Association
Techniques
Ayman F. Habib
Abstract
Download
Full Article
Over the past decade, the need for digital land-related information grew tremendously
to serve the needs of geographic information systems (GIS). Road network
data are of special interest to GIS . Mobile mapping systems (MMS), which
proved to be accurate and efficient data collection devices, can best capture
these data. In this paper, road edges are detected and tracked in a sequence
of image pairs. At successive image pairs, the system discerns the observations
that are believed to originate from the features initialized in the first
stereo pair. Data association algorithms offer the solution to this problem.
The tracking procedure can be carried out for each target (road edge) independently
using the Nearest-Neighbor Standard Algorithm (NNSA). Another approach is
to perform the tracking by considering all targets and observations simultaneously
using the Multiple Hypothesis Approach (MHA). The NNSA is being used
in most photogrammetric applications that require feaÍture tracking
in image sequences. The NNSA performs poorly when several targets compete
for the same observation, especially when we deal with targets that are spatially
close. On the other hand, this problem is accounted for in the MHA (which
has been extensively used in feature tracking by robots equipped with radar).
Data association techniques have the advantage of offering a statistical
measure, which indicates the probability of the matching hypothesis being
the right one. In this paper, tracking road edges in stereo-image sequences
tested both the NNSA and the MHA. As expected, the MHA proved to be
superior to the NNSA. The tracked road edges can be used to update the motion
parameters of the system between the involved image pairs. Also, they can
be used to autonomously navigate the MMS within the tracked road boundaries.
| Top | Home |