Peer-Reviewed Articles
269 Evaluating Object-Based Data Quality Attributes
in the Land Cover Map 2000 of the United Kingdom
Paul Robinson, Peter Fisher, and Geoff Smith
Abstract Download
Full Article
Standards that have been created for the reporting of data
quality in spatial databases focus primarily on database level
metadata, which does not address the issue of varying quality
within a dataset. This issue may be advanced by the inclusion
of metadata at an object level. The Land Cover Map 2000
(LCM2000) is a national database for the UK and contains a
large amount of object-based quality metadata, which is
reviewed. An analysis of uncertainty in the extent of three
land cover types is carried out using a cumulative evidence
method, utilizing a number of existing datasets, each purporting
to represent the extent of the phenomenon in question.
The output of this analysis is compared to the metadata in the
LCM2000, in order to assess the usefulness of the metadata in
understanding attribute accuracy. Results of this comparison
are presented, showing that some of the object-based metadata
gives a useful indication as to the certainty of classification.
277 Automatic Determination
of the Optimum Generic Sensor Model Based on Genetic Algorithm Concepts
Farhad Samadzadegan, Ali Azizi, and Ahmad Abootalebi
Abstract Download
Full Article
Generic sensor models (GSMs) are comprehensive mathematical
models by which different geometric structures of satellite
images could be modeled in order to establish the connection
between image and object spaces. Nevertheless, as they are
mathematical models, rather than physical models, it is
difficult to determine which term and order of GSMs can
provide the best result. Therefore, conventional solutions
need an expert operator to try different terms and orders for
the best solution of GSMs or to find the best trade-off, which
is a complex and time consuming process. Moreover, conventional
solutions for automatic determination of the optimum
GSM parameters are not practically efficient and instead
of going towards the global optimum, frequently get trapped
in some local optima. In this paper we propose a novel
methodology which automatically determines the optimum
GSM’s terms and orders based on genetic algorithm concepts.
Extensive evaluations carried out on a wide range of different
optical satellite images demonstrate the high potentials
of the proposed strategy.
289 Textural Discrimination
of an Invasive Plant, Schinus terebinthifolius, from Low Altitude Aerial
Digital Imagery
Leonard Pearlstine, Kenneth M. Portier, and Scot E. Smith
Abstract Download
Full Article
Schinus terebinthifolius, known as Brazilian pepper, is an
exotic, invasive plant species in Florida that displaces native
plant species and disrupts wildlife habitat. Aerial surveys
typically used to monitor ecosystem change may be augmented
with texture analyses to improve the speed and
consistency with which S. terebinthifolius is detected in the
images. Image processing using high-resolution imagery can
take advantage of high spectral variability in adjacent pixels
of the same cover type by measuring spatial patterns of texture
in neighborhoods of pixels.
Texture features derived from first and second-order statistics and edge components in high-resolution digital color infrared images were tested for their ability to discriminate S. terebinthifolius. Multiple linear logistic regressions found a best subset combination of texture features that consistently identified core areas of S. terebinthifolius. Misclassification of other cover types as S. terebinthifolius was low except where Sabal palmetto was present in the images.
299 Leaf Optical Property
Changes Associated with the Occurrence of Spartina alterniflora Dieback in
Coastal Louisiana Related to Remote Sensing Mapping
Elijah Ramsey III and Amina Rangoonwala
Abstract Download
Full Article
In order to provide a remote sensing solution that would
detect both the initial onset and monitor the early, as well
as, the later stages of impact progression, changes in live
leaf optical properties were compared along transects spanning
impacted coastal Louisiana marsh sites. Green and red
edge reflectance trends generally represented the early stages
and fairly well the later stages of dieback progression, while
blue and red reflectance and absorption trends represented
the later stages of marsh impact that were most closely
related to visible signs of marsh impact. Leaf reflectance in
the near infrared (NIR) was not compatible with visual reflectance
trends and did not co-vary with derived indicators
of leaf water content, and thereby, water stress. Predicted
from reflectance ratios, carotene tended to remain constant
or increase relative to chlorophyll following noted changes
in stressed plants at the two least impacted sites, while the
pigments co-varied at the two most impacted sites. As an
operational solution most amenable for satellite remote
sensing, the NIR/red ratio followed blue and red reflectance
trends while the NIR/green ratio mimicked the green and red
edge reflectance trends indicating impact onset and progression,
as well as, generally portraying blue and red reflectance
trends indicating later stages of impact. The NIR/
green ratio magnitude and range generally increased from
the most to least impacted site providing a convenient
method to detect dieback onset and monitor dieback progression.
This research demonstrated that remote sensing
mapping at these sites could offer a more accurate perception
of dieback severity distribution than offered by determinations
relying on visible indicators of marsh changes.
313 Comparison of Three Algorithms
for Filtering Airborne Lidar Data
Keqi Zhang and Dean Whitman
Abstract Download
Full Article
This paper compares three methods for removing non-ground
measurements from airborne laser scanning data. These
methods, including the elevation threshold with expanding
window (ETEW), maximum local slope (MLS), and progressive
morphological (PM) filters, analyze data points based on
variations of local slope, and elevation. Low and high-relief
data sets with various densities of trees, houses, and sand
dunes were selected to test the filtering methods. The results
show that all three methods can effectively remove most nonground
points in both low-relief urban and high-relief forested
areas. The PM filter generated the best result in coastal barrier
island areas, whereas the other algorithms tended to remove
the tops of steep sand dunes. Each method experienced
various omission or commission errors, depending on the
filtering parameters. Topographic slope is the most sensitive
parameter for the three filtering methods.
325 Semi-Automatic Registration
of Multi-Source Satellite Imagery with Varying Geometric Resolutions
Ayman Habib and Rami Al-Ruzouq
Abstract Download
Full Article
Image registration is concerned with the problem of how to
combine data and/or information from multiple sensors in
order to achieve improved accuracies and better inference
about the environment than could be attained through the
use of a single sensor. Registration of imagery and information
from multiple sources is essential for a variety of applications
in remote sensing, medical diagnosis, computer
vision, and pattern recognition. In general, an image registration
methodology must deal with four issues. First, a
decision has to be made regarding the choice of primitives
for the registration procedure. The second issue is concerned
with establishing the registration transformation function
that mathematically relates geometric attributes of corresponding
primitives. Then, a similarity measure should be
devised to ensure the correspondence of conjugate primitives.
Finally, a matching strategy has to be designed and
implemented as a controlling framework that utilizes the
primitives, the similarity measure, and the transformation
function to solve the registration problem. This paper outlines
a comprehensive investigation and implementation of
the involved issues in a semi-automatic registration procedure
capable of handling multi-source satellite imagery with
varying geometric resolutions.
333 Nested Hyper-Rectangle
Learning Model for Remote Sensing: Land Cover Classification
Li Chen
Abstract Download
Full Article
This study presents an exemplar-based nested hyperrectangle
learning model (NHLM) which is an efficient and
accurate supervised classification model. The proposed
model is based on the concept of seeding training data in
the Euclidean m-space (where m denotes the number of
features) as hyper-rectangles. To express the exceptions,
these hyper-rectangles may be nested inside one another to
an arbitrary depth. The fast and one-shot learning procedures
can adjust weights dynamically when new examples
are added. Furthermore, the “second chance” heuristic is
introduced in NHLM to avoid creating more memory objects
than necessary. NHLM is applied to solving the land cover
classification problem in Taiwan using remote sensed imagery.
The study investigated five land cover classes and
clouds. These six classes were chosen from field investigation
of the study area according to previous study. Therefore,
this paper aims to produce a land cover classification
based on SPOT HRV spectral data. Compared with a standard
back-propagation neural network (BPN), the experimental
results indicate that NHLM provides a powerful tool for
categorizing remote sensing data.