Peer-Reviewed Articles
535 Synchronization of Image Sequences —
A Photogrammetric Method
Karsten Raguse and Christian Heipke
Abstract Download
Full Article
The three-dimensional photogrammetric analysis of image
sequences is a growing field of application. For the analysis
of dynamic processes one important precondition has to be
guaranteed: All cameras have to be synchronized, otherwise
the results are affected by asynchronism. In this article a
new method is presented, which can determine the asynchronism
of an arbitrary number of image sequences. In
contrast to already existing methods, in the new approach
the asynchronism is modeled in object space and then
converted into an interpolation function containing a set of
unknowns for each camera. In this form the asynchronism
is introduced into an extended bundle adjustment, in which
the unknowns are solved simultaneously with the image
orientation parameters and the object coordinates of tie
points. Therefore, the approach has no restrictions with
regard to the number and the set-up of the cameras in the
acquisition network. Furthermore, both the temporal and
spatial analysis step are carried out simultaneously.
We have implemented the suggested method and have run a number of experiments in the context of vehicle impact testing. First, sequences with a frame rate of 1,000 Hz observing an object with a speed of up to 7 m/s and an asynchronism of 0.8 ms were analyzed. The accuracy of the object point determination could be improved by a factor of 10. Then, five sequences of a vehicle impact test with a speed of 15.6 m/s were investigated. Here, errors in the object coordinates of up to 30 mm could be eliminated using the new approach. Given the small tolerances in car development, this improvement in point accuracy is significant.
547 An Emissivity Modulation Method for
Spatial Enhancement of Thermal Satellite Images in Urban Heat Island Analysis
Janet Nichol
Abstract Download
Full Article
This study examines and validates a technique for spatial
enhancement of thermal satellite images for urban heat island
analysis, using a nighttime ASTER satellite image. The technique,
termed Emissivity Modulation, enhances the spatial
resolution while simultaneously correcting the image derived
temperatures for emissivity differences of earth surface
materials. A classified image derived from a higher resolution
visible wavelength sensor is combined with a lower resolution
thermal image in the emissivity correction equation in a
procedure derived from the Stephan Bolzmann law. This has
the effect of simultaneously correcting the image-derived “Brightness Temperature” (Tb) to the true Kinetic Temperature
(Ts), while enhancing the spatial resolution of the
thermal data. Although the method has been used for studies
of the urban heat island, it has not been validated by comparison
with “in situ” derived surface or air temperatures,
and researchers may be discouraged from its use due to the
fact that it creates sharp boundaries in the image. The
emissivity modulated image with 10 m pixel size was found
to be highly correlated with 18 in situ surface and air temperature
measurements and a low Mean Absolute Difference of 1
K was observed between image and in situ surface temperatures.
Lower accuracies were obtained for the Ts and Tb
images at 90 m resolution. The study demonstrates that the
emissivity modulation method can increase accuracy in the
computation of kinetic temperature, improve the relationship
between image values and air temperature, and enable the
observation of microscale temperature patterns.
557 Modification of Pixel-swapping Algorithm
with Initialization fr om a Sub-pixel/pixel
Spatial Attraction Model
Zhangquan Shen, Jiaguo Qi, and Ke Wang
Abstract Download
Full Article
Abstract
Pixel-swapping algorithm is a simple and efficient technique
for sub-pixel mapping (Atkinson, 2001 and 2005). It was
initially applied in shoreline and rural land-cover mapping
but has been expanded to other land-cover mapping.
However, due to its random initializing process, this algorithm
must swap a large number of sub-pixels, and therefore
it is computation intensive. This computing power consumption
intensifies when the scale factor is large. A new,
modified pixel-swapping algorithm (MPS) is presented in this
paper to reduce the computation time, as well as to improve
sub-pixel mapping accuracy. The MPS algorithm replaces the
original random initializing process with a process based on
a sub-pixel/pixel spatial attraction model. The new algorithm
was used to allocate multiple land-covers at the subpixel
level. The results showed that the MPS algorithm
outperformed the original algorithm both in sub-pixel
mapping accuracy and computational time. The improvement
is especially significant in the case of large scale
factors. Furthermore, the MPS is less sensitive to the size of
neighboring sub-pixels and can still result in increased
accuracy even if the size of neighbors is small. The MPS was
also much less time consuming, as it reduced both the
iterations and total amount of swapping needed.
Color Figures (Adobe PDF Format)
569 Closest Spectral Fit for Removing
Clouds and Cloud Shadows
Qingmin Meng, Bruce E. Borders, Chris J. Cieszewski, and Marguerite Madden
Abstract Download
Full Article
Completely cloud-free remotely sensed images are preferred,
but they are not always available. Although the average cloud
coverage for the entire planet is about 40 percent, the removal
of clouds and cloud shadows is rarely studied. To address this
problem, a closest spectral fit method is developed to replace
cloud and cloud-shadow pixels with their most similar nonclouded
pixel values. The objective of this paper is to illustrate
the methodology of the closest spectral fit and test its
performance for removing clouds and cloud shadows in
images. The closest spectral fit procedures are summarized
into six steps, in which two main conceptions, location-based
one-to-one correspondence and spectral-based closest fit, are
defined. The location-based one-to-one correspondence is
applied to identify pixels with the same locations in both base
image and auxiliary images. The spectral-based closest fit is
applied to determine the most similar pixels in an image.
Finally, this closest spectral fit approach is applied to remove
cloud and cloud-shadow pixels and diagnostically checked
using Landsat TM images. Additional examples using Quick-
Bird and MODIS images also indicate the efficiency of the
closest spectral fit for removing cloud pixels.
Color Figures (Adobe PDF Format)
[figure 1.] [figure 2.] [figure 3.] [figure 4.]
577 Land Cover Classification with Multi-Sensor
Fusion of Partly Missing Data
Selim Aksoy, Krzysztof Koperski, Carsten Tusk, and Giovanni Marchisio
Abstract Download
Full Article
We describe a system that uses decision tree-based tools
for seamless acquisition of knowledge for classification of
remotely sensed imagery. We concentrate on three important
problems in this process: information fusion, model
understandability, and handling of missing data. Importance
of multi-sensor information fusion and the use of
decision tree classifiers for such problems have been wellstudied
in the literature. However, these studies have been
limited to the cases where all data sources have a full
coverage for the scene under consideration. Our contribution
in this paper is to show how decision tree classifiers
can be learned with alternative (surrogate) decision nodes
and result in models that are capable of dealing with
missing data during both training and classification to
handle cases where one or more measurements do not exist
for some locations. We present detailed performance
evaluation regarding the effectiveness of these classifiers
for information fusion and feature selection, and study
three different methods for handling missing data in
comparative experiments. The results show that surrogate
decisions incorporated into decision tree classifiers provide
powerful models for fusing information from different data
layers while being robust to missing data.
Color Figures (JPEG Format)
[figure 1.] [figure 2.] [figure 3.] [figure 4.]
595 Spectral Angle Minimization for the Retrieval
of Optically Active Seawater Constituents
from MODIS Data
F. Maselli, L. Massi, M. Pieri, and C. Santini
Abstract Download
Full Article
The application of global algorithms to optical satellite
imagery often fails to correctly assess the concentrations of
seawater constituents (chlorophyll, CHL, suspended sediments,
SS, and yellow substance, YS) in spectrally complex
marine environments. Additional problems may come from
inaccurate radiometric, atmospheric, and geometric corrections
of the remotely sensed imagery. This issue is currently
analyzed using a data set of seawater samples and MODIS
images taken in the Tuscany Sea (Central Italy). The analysis
demonstrates that the mentioned problems mainly introduce
amplitude variations in the measured reflectance. This may
have negative effects on the outcome of inversion algorithms
based on the minimization of conventional spectral errors.
Such effects can be notably reduced by using an error index
derived from the angle between measured and simulated
reflectance vectors, which is insensitive to spectral amplitude
variations. The potential of a classical and the new error
indices is first evaluated by regressing their values against
concentration differences of optically active constituents
found over the available sample pairs. The performance of
the two error indices are then assessed within an inversion
algorithms applied to the same samples. The results obtained
show the potential of the new error index particularly to
improve the estimation of CHL concentration.
607 Multi-temporal RADARSA T-1 and ERS
Backscattering Signatur es of Coastal
Wetlands in Southeaster n Louisiana
Oh-ig Kwoun and Zhong Lu
Abstract Download
Full Article
Using multi-temporal European Remote-sensing Satellites
(ERS-1/-2) and Canadian Radar Satellite (RADARSAT-1) synthetic
aperture radar (SAR) data over the Louisiana coastal
zone, we characterize seasonal variations of radar backscattering
according to vegetation type. Our main findings are as
follows. First, ERS-1/-2 and RADARSAT-1 require careful
radiometric calibration to perform multi-temporal backscattering
analysis for wetland mapping. We use SAR backscattering
signals from cities for the relative calibration. Second,
using seasonally averaged backscattering coefficients from
ERS-1/-2 and RADARSAT-1, we can differentiate most forests
(bottomland and swamp forests) and marshes (freshwater,
intermediate, brackish, and saline marshes) in coastal
wetlands. The student t-test results support the usefulness of
season-averaged backscatter data for classification. Third,
combining SAR backscattering coefficients and an opticalsensor-based normalized difference vegetation index can
provide further insight into vegetation type and enhance the
separation between forests and marshes. Our study demonstrates
that SAR can provide necessary information to
characterize coastal wetlands and monitor their changes.
){kind=link}
){kind=link}
){kind=link}
){kind=link}