Peer-Reviewed Articles
1417 Multi-Sensor System for Airborne Image Capture and Georeferencing
Mohamed M.R. Mostafa and Klaus-Peter Schwarz
Abstract
Download
Full Article
The development and preliminary testing of a fully digital multi-sensor
system for airborne remote sensing and geographic information system
(GIS) applications is described. This system was developed at The
University of Calgary in collaboration with The University of California
at Berkeley, with aircraft and logistics support by HJW Inc., California.
It integrates a medium class inertial navigation system (INS), two
low-cost Global Positioning System (GPS) receivers, and a high resolution
digital camera. During aerial image capture, camera exposure stations
and INS digital records are time-tagged in real time by GPS. The
INS/GPS-derived trajectory parameters describe the rigid body motion
of the carrier aircraft. Thus, they are directly related to the parameters
of exterior orientation. During post-processing, these parameters
are extracted, eliminating the need for ground control for airborne
image acquisition applications. Flight tests were performed over
a part of the university campus at Berkeley, using a strip photography
approach to test the integrated system performance. In this paper,
the concept of direct georeferencing of
digital images without ground control is presented. System calibration results
are then discussed in some detail, and special attention is given to the geometrical
analysis of the system imaging component. An improved imaging system is proposed
and validated by computer simulations. The potential of the new system for
photogrammetric use is then discussed. The major applications of such a system
will be in photo ecometrics; the mapping of utility lines, roads, and pipelines;
and the generation of digital elevation models for engineering applications.
1425 Accuracy Assessment for the U.S. Geological Survey Regional Land-Cover
Mapping Program: New York and New Jersey Region
Zhiliang Zhu, Limin Yang, Stephen V. Stehman, and Raymond L. Czaplewski
Abstract
Download
Full Article
The U.S. Geological Survey, in cooperation with other government and
private organizations, is producing a conterminous U.S. land-cover
map using Landsat Thematic Mapper 30-meter data for the Federal regions
designated by the U.S. Environmental Protection Agency. Accuracy
assessment is to be conducted for each Federal region to estimate
overall and class-specific accuracies. In Region 2, consisting of
New York and New Jersey, the accuracy assessment was completed for
15 land-cover and land-use classes, using interpreted 1:40,000 scale
aerial
photographs as reference data. The methodology used for Region 2 features a
two-stage, geographically stratified approach, with a general sample of all
classes (1,033 sample sites), and a separate sample for rare classes (294 sample
sites). A confidence index was recorded for each land-cover interpretation
on the 1:40,000-scale aerial photography. The estimated overall accuracy for
Region 2 was 63 percent
(standard error 1.4 percent) using all sample sites, and 75.2 percent (standard
error 1.5 percent) using only reference sites with a high-confidence index.
User's and producer's accuracies for the general sample and user's accuracy
for the sample of rare classes, as well as variance for the estimated accuracy
parameters, were also reported. Narrowly defined land-use classes and heterogeneous
conditions of land cover are the major causes of misclassification errors.
Recommendations for modifying the accuracy assessment methodology for use in
the other nine Federal regions are provided.
1439 Digital Land-Use Classification Using Space-Shuttle-Acquired Orbital
Photographs: A Quantitative Comparison with Landsat TM Imagery of
a Coastal Environment, Chanthaburi, Thailand
Edward L. Webb, Ma. Arlene Evangelista, and Julie A. Robinson
Abstract
Download
Full Article
The capability of Space-Shuttle-acquired orbital photography to provide
accurate land-use classification using popular commercial software
and accepted analytical procedures was investigated. The study area
was the coastal region of the Chanthaburi Province, eastern Thailand,
which exhibits a land-use pattern consisting of rice fields, shrimp
farms, plantations/orchards, and patches of healthy and degraded
mangrove habitat. We used a typical image analysis protocol using
ERDAS ImagineTM v8.2 combined with ground referencing, and compared
the classification results using orbital photographs to results of
the same study area using Landsat TM 5 imagery. The orbital photographs
exhibited high spatial resolution, and performed similarly to Landsat
for classification purposes. Accuracy assessments showed 81.3 percent
accuracy of the ground referenced orbital photograph classification,
and 83.3 percent for the Landsat image. Using a GIS overlay, we calculated
71 percent agreement between the two ground referenced image types.
We conclude that, under the appropriate conditions, digitized orbital
photographs can be an excellent source of spatial information for
studies combining images of high spatial and spectral resolution.
In addition to our results, we discuss the benefits and limitations
to using orbital photographs for land-use classification. Orbital
photographs can serve as a low-cost, complementary form of data to
automated
satellite images for assessments of basic habitat parameters.
1451 Land-Use Classification of Remotely Sensed Data Using Kohonen
Self-Organizing Feature Map Neural Networks
C.Y. Ji
Abstract
Download
Full Article
The use of Kohonen Self-Organizing Feature Map (KSOFM, or feature map)
neural networks for land-use/land-cover classification from remotely
sensed data is presented. Different from the traditional multi-layer
neural networks, the KSOFM is a two-layer network that creates class
representation by self-organizing the connection weights from the
input patterns to the output layer. A test of the algorithm is conducted
by classifying a Landsat Thematic Mapper (TM) scene for seven
land-use/land-cover types, benchmarked with the maximum-likelihood method and
the Back Propagation (BP) network. The network outperformes the maximum-likelihood
method for per-pixel classification when four spectral bands are used. A further
increase in classification accuracy is achieved when
neighborhood pixels are incorporated. A similar accuracy is obtained using
the BP networks for classifications both with and without neighborhood information.
The feature map network has the advantage of faster learning but has the drawback
of being a slow classification process. Learning by the feature map is affected
by a number of factors such as the network size, the codebooks partitioning,
the available training samples, and the selection of the learning rate. The
feature map size controls the accuracy at which class borders are formed, and
a large map may be used to obtain accurate class representation. It is concluded
that the feature map
method is a viable alternative for land-use classification of remotely sensed
data.
1461 Water Body Detection and Delineation with Landsat TM Data
Paul Shane Frazier and Kenneth John Page
Abstract
Download
Full Article
The aim of this project was to determine the accuracy of using simple
digital image processing techniques to map riverine water bodies
with Landsat 5 TM data. This paper quantifies the classification
accuracy of single band density slicing of Landsat 5 TM data to delineate
water bodies on riverine floodplains. The results of these analyses
are then compared to a 6-band maximum likelihood classification over
the same area. The water boundaries delineated by each of these digital
classification procedures were compared to water boundaries delineated
from colour aerial photography acquired on the same day as the TM
data. These comparisons show that Landsat TM data can be used to
map water bodies accurately. Density slicing of the single mid-infrared
band 5 proved as successful as multispectral classification achieving
an overall accuracy of 96.9%, a producer's accuracy for water bodies
of 81.7% and a user's accuracy for water bodies of 64.5%.
1469 Predicting the Urbanization of Pine and Mixed Forests in Saint
Tammany Parish, Louisiana
James T. Gunter, Donald G. Hodges, Christopher M. Swalm, and James L. Regens
Abstract
Download
Full Article
St. Tammany Parish, Louisiana, has experienced tremendous urbanization,
resulting in the loss of timberland. This study's objectives were
to develop a model using parish-level data that estimates the probability
of urban development in a pine or mixed forest parcel, and to identify
the parcels most likely to be developed. The geographic data sets
used include satellite imagery from 1981 and 1993, U.S. Census data,
population growth estimates from the St. Tammany Parish Government,
and road coverages. Logistic regression was used to develop a model
that
predicts the probability of urban development. Population density,
distance to the nearest
state or federal highway, distance to the access points to New Orleans, and
distance to the nearest interstate interchange all significantly influenced
the probability that a parcel would be developed. Using population density
predictions for 2003, the model identified the likely development corridor
and the timberlands that would be unavailable for long-term fiber production.
1477 Semi-Automated Object Measurement Using Multiple-Image Matching
from Mobile Mapping Image Sequences
C. Vincent Tao
Abstract
Download
Full Article
Mobile mapping systems offer rapid acquisition of georeferenced image
sequences. With the vast amounts of image sequences collected by
these systems, efficient and accurateinformation extraction from
image sequences becomes a critical issue. It affects the operational
cost of the application of the mobile mapping technology. The goal
of this research is to develop an efficient and robust approach for
accurate object measurement from mobile mapping image sequences of
road corridors. The paper describes a semi-automated object measurement
approach based on the use of a multiple-image matching strategy.
Once an object point in one image is
measured manually, the corresponding points in consecutive image pairs can
be determined automatically, and then, the 3D coordinates of the object
point can be calculated by photogrammetric intersection using multiple corresponding
points. Based on the test results, the performance of this approach was operationally
satisfactory in terms of reliability, accuracy, and efficiency. The effective
use of stereoscopic and sequential image information, known image geo-referencing
parameters, and multiple-baseline geometry is a key to the development of this
semi-automated object measurement approach.
| Top | Home |