Peer-Reviewed Articles
685 Redefining the Paradigm of Modern Mobile Mapping: An
Automated High-Precision Road Centerline Mapping System
Charles Toth and Dorota Grejner-Brzezinska
Abstract Download
Full Article
A high-precision land-based integrated mapping system has been developed at
The Ohio State University to support road centerline mapping operations at
the Ohio Department of Transportation District 1 Office. The system represents
a transition from the traditional mobile mapping paradigm towards a highly
automated and autonomous design following the trends of modern geoinformatics.
The two key components of the custom-designed system are a high-precision
integrated GPS/INS navigation system and a fully digital and automated imaging
subsystem. The van-based mapping system was designed to deliver the road
centerline positions at subdecimeter accuracy in a highly automated manner
with limited human interaction in near real time. The paper presents the
system's concept and design, followed by an individual performance evaluation
of the navigation and imaging components; and finally road test results,
representing an operational environment, are also reported.
Please see these links for the color figures:
Figure 1. Figure
2. Figure 3. Figure
4. Figure 5. Figure
6. Figure 7. Figure
8. Figure 9. Figure
10. Figure 11. Figure
12.
695 DTM Generation from Ikonos In-Track Stereo
Images Using a 3D Physical Model
Thierry Toutin
Abstract Download
Full Article
A digital elevation model (DEM) extracted from Ikonos in-track stereo images
using a 3D physical model developed at the Canada Centre for Remote Sensing,
Natural Resources Canada was evaluated. First, the stereo photogrammetric
bundle adjustment was set up with about ten accurate ground control points.
The DEM was then generated using an area-based multiscale image matching
method and 3D semi automatic editing tools and then compared to lidar elevation
data with a 0.2-m accuracy. Because the DEM is, in fact, a digital terrain
surface model where the height of land cover (trees, houses) is included,
the accuracy varies depending on land cover types. Using 3D visual classification
of the stereo Ikonos images, different classes (forests, residential, bare
soil, lakes) were generated to take into account the height of the surface
(natural and human-made) in the accuracy evaluation. An elevation linear
error with 68 percent confidence level (LE68) of 1.5 m was obtained for bare
surfaces while an LE68 of 6.4 m was achieved over the full area. Five-meter
contour lines could thus be derived, compliant with the highest topographic
standard. Better results could thus be expected when using stereo-images
acquired in the summertime. On the other hand, an LE68 of 2.5 m to 6.6 m
was obtained depending on the land-cover type and its surface height. For
residential areas, the surface height did not affect the errors very much
(2.5-m LE68) when compared to bare surface results because one-and two-story
houses were sparse in the test area. Because the images were unfortunately
acquired in wintertime and the lidar data in summertime, elevation errors
(LE68 and bias) also depended on the type of forest (deciduous, coniferous,
mixed, sparse). An evaluation based on terrain slope and azimuth showed that
the DEM error was linearly correlated with slope and that elevations on sun-facing
slopes were 1-m more accurate than elevations on slopes facing away from
the sun.
703 Mapping Coastal Vegetation Using an Expert
System and Hyperspectral Imagery
K.S. Schmidt, A.K. Skidmore, E.H. Kloosterman, H. Van Oosten, L. Kumar, and
J.A.M. Janssen
Abstract Download
Full Article
Mapping and monitoring saltmarshes in the Netherlands are important activities
of the Ministry of Public Works (Rijkswaterstaat). The Survey Department
(Meetkundige Dienst) produces vegetation maps using aerial photographs. However,
it is a time-consuming and expensive activity. The accuracy of the conventional
vegetation map derived using aerial photograph interpretation (API) is estimated
to be around 43 percent. In this study, an alternative method is demonstrated
that uses an expert system to combine airborne hyperspectral imagery with
terrain data derived from radar altimetry. The accuracy of the vegetation
map generated by the expert system increased to 66 percent. When hyperspectral
imagery alone was used to classify coastal wetlands, an accuracy of 40 percent
was achieved-comparable to the accuracy of the API-derived vegetation map.
An analysis of the efficiency of the proposed expert system showed that the
speed of map production is increased by using the new method. This means
that digital image classification using the expert system is an objective
and repeatable method superior to the conventional API method.
717 Using Remote Sensing to Assess Stand Loss
and Defoliation in Maize
Bruce J. Erickson, Chris J. Johannsen, James J. Vorst, and Larry L. Biehl
Abstract Download
Full Article
Assessing hail and wind damage to crops is a difficult, labor intensive task.
A quick and accurate method of determining losses could lead to better crop
management decisions, more accurate insurance claim adjustment, and reduced
expenses for the crop hail insurance industry. Radiometric data were collected
in 1997, 1998, and 1999 in Indiana and Nebraska from field plots of maize,
Zea mays L., subjected to varying levels of damage. Incremental differences
in plant damage resulted in incremental differences in spectral responses.
The red and near-infrared spectral bands provided the most discrimination
among levels of damage. Classification of remotely sensed images by damage
level was performed by extrapolating spectral information from areas where
damage levels were known to adjacent unknown areas of damage. Depending on
location, sensor, and date of data collection, it was possible to classify
the degree of early-season stand loss at accuracies of 48 to 100 percent.
For leaf loss during the late vegetative stages, it was possible to classify
the degree of leaf loss at accuracies of 81 to 100 percent and, for leaf
loss during the early reproductive stages, it was possible to classify damage
at accuracies of 71 to 98 percent. These results indicate that remote sensing
could be used to improve the accuracy of estimating crop damage as long as
adequate ground reference for different levels of crop damage exists.
723 Comparison of Land-Cover Classification Methods
in the Brazilian Amazon Basin
Dengsheng Lu, Paul Mausel, Mateus Batistella, and Emilio Moran
Abstract Download
Full Article
Four distinctly different classifiers were used to analyze multi-spectral data.
Which of these classifiers is most suitable for a specific study area is
not always clear. This paper provides a comparison of minimum-distance classifier
(MDC), maximum likelihood classifier (MLC), extraction and classification
of homogeneous objects (ECHO), and decision-tree classifier based on linear
spectral mixture analysis (DTC-LSMA). Each of the classifiers used both Landsat
Thematic Mapper data and identical field-based training sample datasets in
a western Brazilian Amazon study area. Seven land-cover classes; mature forest,
advanced secondary succession, initial secondary succession, pasture lands,
agricultural lands, bare lands, and water-were classified. Classification
results indicate that the DTC-LSMA and ECHO classifiers were more accurate
than were the MDC and MLC. The overall accuracy of the DTC-LSMA approach
was 86 percent with a 0.82 kappa coefficient and ECHO had an accuracy of
83 percent with a 0.79 kappa coefficient. The accuracy of the other classifiers
ranged from 77 to 80 percent with kappa coefficients from 0.72 to 0.75.
733 CoLaPS: An Integrated System Linking Production
and Utilization of Land-Cover Information Derived from Landsat Data
Bert Guindon and Ying Zhang
Abstract Download
Full Article
A system, CoLaPS (Composite Land Processing System), is described for the production
and analysis of land-cover information derived from Landsat satellite images.
A key aspect of the design is the seamless linkage between its production
and user subsystems. This element is based upon a number of postulates. First,
for more efficient and diverse usage of remote sensing land-cover products,
users require access to a broader range of datasets and functionality, many
of which traditionally have resided in the production arena. Availability
of these can result in improved (1) visual interpretation, (2) product enhancement
by harnessing user expertise and regional ancillary information, and (3)
higher level landscape characterizations. A key challenge facing land-cover
producers is the need for detailed accuracy characterization of their products.
CoLaPS utilizes classification consistency as an accuracy surrogate, leading
to measures of classification confidence at the pixel level. This greatly
enhances a user's ability to detect, for example, real thematic change between
multitemporal land-cover products.
743 A Critical Evaluation of the Normalized Error
Matrix in Map Accurcy Assessment
Stephen V. Stehman
Abstract Download
Full Article
Normalizing an error matrix is a commonly recommended analysis of map accuracy
data. Theoretical and empirical results demonstrate that the traditional
practice of normalizing an error matrix to uniform homogeneous marginal proportions
produces biased and imprecise accuracy estimates, with the bias most prominent
for user's and producer's accuracies. When used to compare maps, normalizing
to uniform homogeneous marginal proportions evaluates an unrealistic hypothetical
scenario in which all classes are assumed present in equal proportions for
both the map and reference condition, and the map exactly reproduces the
reference area proportions for each class. The marginal proportions of such
normalized error matrices do not reflect realistic area distributions of
either the map or true condition. For both descriptive and comparative objectives,
the advantages typically claimed for normalizing error matrices are far outweighed
by the estimation and interpretation difficulties created by this practice.
| Top | Home |