Segmentation appears to be an attractive preprocessing state when performing point target detecti... more Segmentation appears to be an attractive preprocessing state when performing point target detection in hyperspectral data. Nevertheless, the literature contains examples of both successful and unsuccessful segmentation. Using simulations and several new analytical tools, we propose to derive guidelines when segmentation would be useful and when it would be superfluous. A real dataset with parameters when segmentation is worthwhile and when not is given.
This article is an open access article distributed under the terms and conditions of the Creative... more This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY
EGU General Assembly Conference Abstracts, Apr 1, 2019
Wind streaks are common aeolian features observed on planetary surfaces. They occur by the thousa... more Wind streaks are common aeolian features observed on planetary surfaces. They occur by the thousands on Mars, Venus and Earth and it has been suggested that they form also on the moons Titan and Triton. Due to their global distribution, compared to other aeolian features, wind streaks serve as a primary indicator for the prevailing wind direction near the surface over time. The climatic perspective of wind streaks in the planetary context has been supported by recent studies on Earth, their interpretation in remote sensing images is still lacking. The geomorphological properties of wind streaks are diverse and may involve changes in surface composition, thickness, grain size or the presence of aeolian bedforms compared to their surroundings. The appearance of wind streaks in optical or radar images is significantly different. Optical images are mostly affected by the mineralogy of the surface, whereas, radar is sensitive to the surface roughness and the viewing geometry. Because Venus is constantly shrouded with clouds, its surface was imaged solely by radar. Mars, on the other hand, has been imaged only by optical sensors (visible and infrared spectrum). Therefore, the observations of Martian and Venusian wind streaks are not comparable. Currently, Earth is the only planet where wind streaks were imaged by both optical and radar sensors and a comprehensive data set can be created. This study presents a wide-ranging comparison of remote sensing means, using various sensors, both optical and radar. The purpose of this study is to examine the effects imposed by the sensor type on wind streak identification and interpretation. Six case studies of Earth wind streaks are demonstrated which were imaged by both radar and optical sensors. Results indicate that wind streak identification is constrained to a specific combination of sensor settings fit for the local surface properties. Bright and dark (reflectance and backscatter) are relative terms and should be used with caution. Only half of the optical-visible wind streaks are also radar-visible, but all wind streaks observed in radar images were observed in optical images. The results suggest that the Venusian wind streak population was not revealed entirely and more wind streaks are waiting to be discovered. In addition, data collected during field investigations from five sites in Mojave Desert, California, USA supports the results obtained by remote sensing. Field data includes spectral measurement, GPS, documentation of the environment conditions and soil sampling for further laboratory analyses. Preliminary results from x-ray diffraction analysis and spectral signatures of both the streak and surroundings imply on little mineralogical difference in four case studies. Spectral differences between the streaks and their surroundings correspond to what is observed and measured in imagery.
In this article, a new algorithm for point target detection based on nonnegative matrix factoriza... more In this article, a new algorithm for point target detection based on nonnegative matrix factorization (NMF) is presented. The basic concept is outlined and the proposed algorithm is compared to previously known matched filter (MF) based algorithms. This comparison indicates improvement in detection rates with the new algorithm.
In this paper, we present a variation on the LRX (Local RX) algorithm for detecting anomalies in ... more In this paper, we present a variation on the LRX (Local RX) algorithm for detecting anomalies in multi-temporal images. Our algorithm assigns a relative weight to the Mahalanobis distance according to the number of times it appears in an image. Standard transitions between pixels are therefore not viewed as anomalous; unusual transitions are assigned proportionally higher weights. Experimental results using our proposed algorithm vs previous algorithms on multitemporal datasets show a significant improvement.
Sand seas and dune bedforms are climatic chronicles that have a wide range of scientific meanings... more Sand seas and dune bedforms are climatic chronicles that have a wide range of scientific meanings. The Karakum and Kyzylkum sand seas in Central Asia span ∼500,000 km2, among the largest sand seas on Earth; yet, very little is known about their emergence and stabilization. Further, the internationally accessible literature lacks comprehensive studies on the interactions between contemporary climate and dune morphodynamics for this region. In this study, we applied the conventional approach for examining wind data—Fryberger’s drift equations and the latest Copernicus ERA5 wind reanalysis model—to demonstrate that the Karakum and Kyzylkum deserts had (and continue to have) low-energy environments during 1950–2019. Time-series multidimensional raster analysis and data mining indicated that 93% of surveyed desert areas were subjected to low-energy wind conditions during the last 70 years. The decrease in wind energy was further confirmed using the partitioning around medoids (PAM) unsup...
2019 IEEE International Conference on Microwaves, Antennas, Communications and Electronic Systems (COMCAS)
One of most common algorithms for anomaly detection in multi-dimensional imagery is the Reed - Xi... more One of most common algorithms for anomaly detection in multi-dimensional imagery is the Reed - Xiaoli (RX) algorithm; it gives each pixel a score that defines its likelihood to be an anomaly. We have implemented a new algorithm which uses both RX and the Non-Negative Matrix Factorization (NNMF) learning algorithm in order to pick an adaptive threshold for detection; we have applied it to Synthetic Aperture Radar (SAR) data. The NNMF approach is defined as a minimization problem which approximates the given data by extracting its main trends. By comparing the original data to the reduced data, we can divide the image anomalies into two different groups, where one group contains the anomalies which are part of the image main trends and the second group contains the anomalies of the sub trends. With this division, we can pick an adaptive threshold for each of the groups according to its unique characteristics.
<p>Much geomorphological research has been conducted on the origin and evolution of... more <p>Much geomorphological research has been conducted on the origin and evolution of sand seas, some of which focused on categorizing dune types, some determining past climatic conditions using Optically Stimulated Luminescence (OSL) dating, and some, understanding dune patterns and stability in relation to the current climate. Sand seas are, in essence, climate archives. Even though the central Asian sand seas, Karakum and Kyzylkum, occupy more than 450.000 km2, relatively little climatic research has been conducted on the formation and evolution of these aeolian deserts. The meteorological data (current & historical) for these regions is scarce and incomplete. Moreover, the area is not extensively recognized in the worldwide available aeolian literature, which has little information about its climatic performances. Using the common method for the examination of wind data—Fyberger’s Drift Erosion equations (1979) and the most recent Copernicus ERA5 wind reanalysis model conducted on 5081 control grid points, we show that both the Karakum and Kyzylkum Deserts are and have been low-energy environments for the statistical period 1950–2019. The data set was transferred into the form of a space-time cube, a multidimensionally-orientated raster, enabling time-series analysis and data mining. Such analysis has revealed that 93% of the desert was under a low-energy wind setting during the past 70 years. The directionality index shows that erosive winds didn’t deviate extensively from their primal directions as observed from the same statistical period. Such performance favors vegetation growth and biogenic crust development, thus, dune stabilization. Alongside, a similar type of procedure was performed for the wind data obtained from NOAA weather stations that are at close distance or within the study area. Both data sets were mutually analyzed with respect to the Resultant Drift Direction (RDD) calculated for the period 2010-2019 and all available portions of time. By comparing dune orientation and characteristics of their current morphology to the wind directions, the results indicate that the morphology of active aeolian features, (e.g. complex barchanoid ridges, star and seif dunes) are in high correlation with current winds and their erosive powers. Simultaneously, stable aeolian formations that overlay most of the observed area (determined as different formations of Vegetated Linear Dunes - VLDs) show disassociation between the dune's orientation and the wind power (RDP), direction (RDD) and variability (UDI), confirming that most of the vegetated aeolian formations were not developed, or presently reworked, by contemporary winds. An analysis of wind parameters in both active and stable aeolian morphologies in these deserts has permitted the discovery of present and paleo-wind interactions, as well as putative paleo-environmental climatic conditions.</p>
Image and Signal Processing for Remote Sensing XXIII, 2017
During the past years, several compressive spectral imaging techniques were developed. With these... more During the past years, several compressive spectral imaging techniques were developed. With these techniques, an optically compressed version of the spectral datacube is captured. Consequently, the information about the object and targets is captured in a lower dimensional space. A question that rises is whether the reduction of the captured space affects the target detection performance. The answer to this question depends on the compressive spectral imaging technique employed. In most compressive spectral imaging techniques, the target detection performance is deteriorated. We show that our recently introduced technique, dubbed Compressive Sensing Miniature Ultra-Spectral Imaging (CSMUSI), yields similar target detection and false detection rates to that of conventional hyperspectral cameras.
Segmentation appears to be an attractive preprocessing state when performing point target detecti... more Segmentation appears to be an attractive preprocessing state when performing point target detection in hyperspectral data. Nevertheless, the literature contains examples of both successful and unsuccessful segmentation. Using simulations and several new analytical tools, we propose to derive guidelines when segmentation would be useful and when it would be superfluous. A real dataset with parameters when segmentation is worthwhile and when not is given.
This article is an open access article distributed under the terms and conditions of the Creative... more This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY
EGU General Assembly Conference Abstracts, Apr 1, 2019
Wind streaks are common aeolian features observed on planetary surfaces. They occur by the thousa... more Wind streaks are common aeolian features observed on planetary surfaces. They occur by the thousands on Mars, Venus and Earth and it has been suggested that they form also on the moons Titan and Triton. Due to their global distribution, compared to other aeolian features, wind streaks serve as a primary indicator for the prevailing wind direction near the surface over time. The climatic perspective of wind streaks in the planetary context has been supported by recent studies on Earth, their interpretation in remote sensing images is still lacking. The geomorphological properties of wind streaks are diverse and may involve changes in surface composition, thickness, grain size or the presence of aeolian bedforms compared to their surroundings. The appearance of wind streaks in optical or radar images is significantly different. Optical images are mostly affected by the mineralogy of the surface, whereas, radar is sensitive to the surface roughness and the viewing geometry. Because Venus is constantly shrouded with clouds, its surface was imaged solely by radar. Mars, on the other hand, has been imaged only by optical sensors (visible and infrared spectrum). Therefore, the observations of Martian and Venusian wind streaks are not comparable. Currently, Earth is the only planet where wind streaks were imaged by both optical and radar sensors and a comprehensive data set can be created. This study presents a wide-ranging comparison of remote sensing means, using various sensors, both optical and radar. The purpose of this study is to examine the effects imposed by the sensor type on wind streak identification and interpretation. Six case studies of Earth wind streaks are demonstrated which were imaged by both radar and optical sensors. Results indicate that wind streak identification is constrained to a specific combination of sensor settings fit for the local surface properties. Bright and dark (reflectance and backscatter) are relative terms and should be used with caution. Only half of the optical-visible wind streaks are also radar-visible, but all wind streaks observed in radar images were observed in optical images. The results suggest that the Venusian wind streak population was not revealed entirely and more wind streaks are waiting to be discovered. In addition, data collected during field investigations from five sites in Mojave Desert, California, USA supports the results obtained by remote sensing. Field data includes spectral measurement, GPS, documentation of the environment conditions and soil sampling for further laboratory analyses. Preliminary results from x-ray diffraction analysis and spectral signatures of both the streak and surroundings imply on little mineralogical difference in four case studies. Spectral differences between the streaks and their surroundings correspond to what is observed and measured in imagery.
In this article, a new algorithm for point target detection based on nonnegative matrix factoriza... more In this article, a new algorithm for point target detection based on nonnegative matrix factorization (NMF) is presented. The basic concept is outlined and the proposed algorithm is compared to previously known matched filter (MF) based algorithms. This comparison indicates improvement in detection rates with the new algorithm.
In this paper, we present a variation on the LRX (Local RX) algorithm for detecting anomalies in ... more In this paper, we present a variation on the LRX (Local RX) algorithm for detecting anomalies in multi-temporal images. Our algorithm assigns a relative weight to the Mahalanobis distance according to the number of times it appears in an image. Standard transitions between pixels are therefore not viewed as anomalous; unusual transitions are assigned proportionally higher weights. Experimental results using our proposed algorithm vs previous algorithms on multitemporal datasets show a significant improvement.
Sand seas and dune bedforms are climatic chronicles that have a wide range of scientific meanings... more Sand seas and dune bedforms are climatic chronicles that have a wide range of scientific meanings. The Karakum and Kyzylkum sand seas in Central Asia span ∼500,000 km2, among the largest sand seas on Earth; yet, very little is known about their emergence and stabilization. Further, the internationally accessible literature lacks comprehensive studies on the interactions between contemporary climate and dune morphodynamics for this region. In this study, we applied the conventional approach for examining wind data—Fryberger’s drift equations and the latest Copernicus ERA5 wind reanalysis model—to demonstrate that the Karakum and Kyzylkum deserts had (and continue to have) low-energy environments during 1950–2019. Time-series multidimensional raster analysis and data mining indicated that 93% of surveyed desert areas were subjected to low-energy wind conditions during the last 70 years. The decrease in wind energy was further confirmed using the partitioning around medoids (PAM) unsup...
2019 IEEE International Conference on Microwaves, Antennas, Communications and Electronic Systems (COMCAS)
One of most common algorithms for anomaly detection in multi-dimensional imagery is the Reed - Xi... more One of most common algorithms for anomaly detection in multi-dimensional imagery is the Reed - Xiaoli (RX) algorithm; it gives each pixel a score that defines its likelihood to be an anomaly. We have implemented a new algorithm which uses both RX and the Non-Negative Matrix Factorization (NNMF) learning algorithm in order to pick an adaptive threshold for detection; we have applied it to Synthetic Aperture Radar (SAR) data. The NNMF approach is defined as a minimization problem which approximates the given data by extracting its main trends. By comparing the original data to the reduced data, we can divide the image anomalies into two different groups, where one group contains the anomalies which are part of the image main trends and the second group contains the anomalies of the sub trends. With this division, we can pick an adaptive threshold for each of the groups according to its unique characteristics.
<p>Much geomorphological research has been conducted on the origin and evolution of... more <p>Much geomorphological research has been conducted on the origin and evolution of sand seas, some of which focused on categorizing dune types, some determining past climatic conditions using Optically Stimulated Luminescence (OSL) dating, and some, understanding dune patterns and stability in relation to the current climate. Sand seas are, in essence, climate archives. Even though the central Asian sand seas, Karakum and Kyzylkum, occupy more than 450.000 km2, relatively little climatic research has been conducted on the formation and evolution of these aeolian deserts. The meteorological data (current & historical) for these regions is scarce and incomplete. Moreover, the area is not extensively recognized in the worldwide available aeolian literature, which has little information about its climatic performances. Using the common method for the examination of wind data—Fyberger’s Drift Erosion equations (1979) and the most recent Copernicus ERA5 wind reanalysis model conducted on 5081 control grid points, we show that both the Karakum and Kyzylkum Deserts are and have been low-energy environments for the statistical period 1950–2019. The data set was transferred into the form of a space-time cube, a multidimensionally-orientated raster, enabling time-series analysis and data mining. Such analysis has revealed that 93% of the desert was under a low-energy wind setting during the past 70 years. The directionality index shows that erosive winds didn’t deviate extensively from their primal directions as observed from the same statistical period. Such performance favors vegetation growth and biogenic crust development, thus, dune stabilization. Alongside, a similar type of procedure was performed for the wind data obtained from NOAA weather stations that are at close distance or within the study area. Both data sets were mutually analyzed with respect to the Resultant Drift Direction (RDD) calculated for the period 2010-2019 and all available portions of time. By comparing dune orientation and characteristics of their current morphology to the wind directions, the results indicate that the morphology of active aeolian features, (e.g. complex barchanoid ridges, star and seif dunes) are in high correlation with current winds and their erosive powers. Simultaneously, stable aeolian formations that overlay most of the observed area (determined as different formations of Vegetated Linear Dunes - VLDs) show disassociation between the dune's orientation and the wind power (RDP), direction (RDD) and variability (UDI), confirming that most of the vegetated aeolian formations were not developed, or presently reworked, by contemporary winds. An analysis of wind parameters in both active and stable aeolian morphologies in these deserts has permitted the discovery of present and paleo-wind interactions, as well as putative paleo-environmental climatic conditions.</p>
Image and Signal Processing for Remote Sensing XXIII, 2017
During the past years, several compressive spectral imaging techniques were developed. With these... more During the past years, several compressive spectral imaging techniques were developed. With these techniques, an optically compressed version of the spectral datacube is captured. Consequently, the information about the object and targets is captured in a lower dimensional space. A question that rises is whether the reduction of the captured space affects the target detection performance. The answer to this question depends on the compressive spectral imaging technique employed. In most compressive spectral imaging techniques, the target detection performance is deteriorated. We show that our recently introduced technique, dubbed Compressive Sensing Miniature Ultra-Spectral Imaging (CSMUSI), yields similar target detection and false detection rates to that of conventional hyperspectral cameras.
Uploads
Papers by Dan G Blumberg