Paul Gader - Academia.edu (original) (raw)
Papers by Paul Gader
IEEE Transactions on Acoustics, Speech, and Signal Processing, 1989
ABSTRACT
Linear Algebra and its Applications, 1988
CVGIP: Image Understanding, 1991
... We use the notation of the Ritter image algebra. We consider the important case of translatio... more ... We use the notation of the Ritter image algebra. We consider the important case of translation invariant, rec tangular templates defined on a twodimensional coordi nate set. ... 6. TR Crimmins and WM Brown, Image algebra and automatic shape recognition, IEEE Trans. ...
Detection and Remediation Technologies for Mines and Minelike Targets V, 2000
ABSTRACT
Detection and Remediation Technologies for Mines and Minelike Targets IX, 2004
ABSTRACT
SPIE Proceedings, 1996
The Choquet fuzzy integral provides a useful mechanism for evidence aggregation. It is a flexible... more The Choquet fuzzy integral provides a useful mechanism for evidence aggregation. It is a flexible method which can represent weighted averages, medians, order statistics, and many other information aggregation mechanisms. In this paper, two applications are described to handwritten word recognition: as a match function in a dynamic programming based classifier and as a method for fusing the results from multiple word recognition algorithms. In the first case, the results are compared with traditional methods. In the second case, the results are compared with neural network and Borda count approaches.
Radio Science, 2004
ABSTRACT The utility of acoustic-to-seismic coupling systems for landmine detection has been clea... more ABSTRACT The utility of acoustic-to-seismic coupling systems for landmine detection has been clearly established. In this approach, laser Doppler vibrometers (LDV) are used to measure the different responses to acoustic excitation in ground regions with and without buried landmines. Currently, for most applications, only the magnitude of the surface velocity is investigated and used to construct recognition algorithms. Recently, we introduced phase-based features in the classification scheme, significantly lowering false alarm rates at given detection probabilities. In this paper, we present modeling equations that explain the phase features for ground areas both with and without buried landmines from the perspective of harmonic oscillator models. We also describe the image processing techniques applied to velocity data collected in the time domain with a moving LDV array. The observed signatures are also compared with the prediction of the models described. We also construct classifiers with only magnitude information and both magnitude and phase information for this time-domain data set. Classification results indicate that we can combine magnitude and phase features to improve the detection of buried mines while reducing false alarms. We also find that using phase information improves the distinction between ground regions with buried landmines or man-made clutter objects.
Linear Algebra and its Applications, 1990
The Journal of the Acoustical Society of America, 2004
ABSTRACT Acoustic/seismic systems demonstrate significant potential for reliable detection of lan... more ABSTRACT Acoustic/seismic systems demonstrate significant potential for reliable detection of landmines, particularly when coupled with ground-penetrating radar (GPR). Acoustic/seismic systems provide complementary information to GPR. One difficulty in developing detection algorithms for acoustic/seismic systems is that the frequency bands at which information occurs vary according to factors such as mine depth and mechanical properties of the soil, which are not known during mine detection. Ordered weighted-averaging (OWA) systems offer the potential for robust detection of landmines under different conditions because they naturally provide the capability to process the best collection of frequency bands in a fashion that can be optimized. In this paper, a processing framework is presented that combines OWA operators for feature analysis with decision making. This framework can be optimized using acoustic/seismic signals as well as acoustic/seismic signals combined with GPR. In the latter case, the optimization can help to provide the appropriate mechanism for combining the complementary information from both sensors. Results are presented using real data from both mines and other buried objects collected at an outdoor test site.
Journal of Electronic Imaging, 2000
Journal of Electronic Imaging, 1996
ABSTRACT The Choquet fuzzy integral is applied to handwritten word recognition. A handwritten wor... more ABSTRACT The Choquet fuzzy integral is applied to handwritten word recognition. A handwritten word recognition system is described. The word recognition system assigns a recognition confidence value to each string in a lexicon of candidate strings. The system uses a lexicon-driven approach that integrates segmentation and recognition via dynamic programming matching. The dynamic programming matcher finds a segmentation of the word image for each string in the lexicon. The traditional match score between a segmentation and a string is an average. In this paper, fuzzy integrals are used instead of an average. Experimental results demonstrate the utility of this approach. A surprising result is obtained that indicates a simple choice of fuzzy integral works better than a more complex choice.
ABSTRACT The HSTAMIDS hand-held mine detection system can be operated in scan mode and investigat... more ABSTRACT The HSTAMIDS hand-held mine detection system can be operated in scan mode and investigation mode. In investigation mode, it is assumed that the system has produced an initial alarm at a location and the operator investigates the region around the location intensively. This mode of operation provides a different set of assumptions and requirements for detection algorithms operating within the unit. This paper discusses how the assumptions and requirements change and describes experimental laboratory results (using data collected in the field) in which significant improvements in Probability of Detection vs. False Alarm Rates were achieved. Two methods for region processing are discussed. Both methods transfer the frequency domain data acquired by the HSTAMIDS system into the time domain. One method is based on Size-Contrast Filtering and the other is based on robustly measuring the variation along the region corresponding to the potential mine location. These new methods are compared to the Correlation Detector, an online algorithm for detecting mines with the HSTAMIDS GPR which has been described in previous papers. Particular attention is given to deep Anti-Tank mines, which are difficult because they too far from the metal detector to produce a metal detector response. In this case, a factor of 6 reduction in FAR is obtained at 100% Probability of Detection.
ABSTRACT For explosive detection purposes, it is assumed that the per- son preparing or carrying ... more ABSTRACT For explosive detection purposes, it is assumed that the per- son preparing or carrying the explosive will inadvertently contaminate him/herself or the exterior of the package. To detect such traces of explosive materials, we show the use of differential reflectometry (DR) as an alternative system to the existing techniques. With DR, explosives show char- acteristic behaviours at specific wavelengths, for example, spectra of TNT shows a sudden decrease at 420 nm. To detect these behaviours, principle component analysis was performed to reduce the dimensionality of the data, and a support vector machine classifier was trained to identify TNT. With a 10-fold classification on 10000 non-TNT and 1935 TNT pixels, we achieved 0.3% false alarm rate at 75% true positive rate. In this study, we outline the operation of the DR system, show the unique signatures of explosives when viewed with DR, and report the detection rates with support vector machine classifiers.
SPIE Proceedings, 2007
We propose a general method for detecting landmine signatures in vehicle mounted ground penetrati... more We propose a general method for detecting landmine signatures in vehicle mounted ground penetrating radar (GPR) using discrete hidden Markov models and Gabor wavelet features. Observation vectors are constructed based on the expansion of the signature's B-scan using a bank of ...
Context-based unmixing has been studied by several re-searchers. Recent techniques, such as piece... more Context-based unmixing has been studied by several re-searchers. Recent techniques, such as piece-wise convex unmixing using fuzzy and possibilistic clustering or Bayesian methods proposed in [11] attempt to form contexts via clus-tering. It is assumed that the linear mixing model applies to each cluster (context) and endmembers and abundances are found for each cluster. As the clusters are spatially coher-ent, hyperspectral image segmentation can significantly aid unmixing approaches that perform cluster specific estimation of endmembers. In this work, we integrate a graph-cuts seg-mentation algorithm with piece-wise convex unmixing. This is compared to fuzzy clustering (FCM) with results obtained on two datasets. The results demonstrate that the integrated approach achieves better segmentation and more precise end-member identification (in terms of comparisons with known ground truth).
![Research paper thumbnail of WHISPERS 2013: 5th Workshop on Hyperspectral Image and Signal Processing ? Evolution in Remote Sensing [Conference Reports]](https://attachments.academia-assets.com/87681228/thumbnails/1.jpg)
](IEEE Geoscience and Remote Sensing Magazine, 2013
SPIE Proceedings, 2005
ABSTRACT Identifying unique patterns of energy in ground penetrating radar images plays an import... more ABSTRACT Identifying unique patterns of energy in ground penetrating radar images plays an important role in landmine/clutter discrimination. Many different geometric features, including size and the distribution of energy values in a radar image, can be exploited in mine detection and discrimination. The granulometry of a random set (image), computed by measuring the integral of a sequence of closings with elements from a family of increasing homothetic shapes, yields a size distribution that can be used for texture analysis or object detection and discrimination. An important complement to granulometries for discrimination is the compactometry, a feature we have identified that is computed by measuring the integral of a sequence of increasing concentric homothetic subsets of a random set. The compactometry yields a characterization of the concentration of the density distribution of the random set at a given point. This paper investigates the properties of compactometry and its derivative, the density spectrum, and demonstrates how they can be used together with granulometry to address the problem of landmine/clutter discrimination using ground-penetrating radar sensors.
IEEE Transactions on Acoustics, Speech, and Signal Processing, 1989
ABSTRACT
Linear Algebra and its Applications, 1988
CVGIP: Image Understanding, 1991
... We use the notation of the Ritter image algebra. We consider the important case of translatio... more ... We use the notation of the Ritter image algebra. We consider the important case of translation invariant, rec tangular templates defined on a twodimensional coordi nate set. ... 6. TR Crimmins and WM Brown, Image algebra and automatic shape recognition, IEEE Trans. ...
Detection and Remediation Technologies for Mines and Minelike Targets V, 2000
ABSTRACT
Detection and Remediation Technologies for Mines and Minelike Targets IX, 2004
ABSTRACT
SPIE Proceedings, 1996
The Choquet fuzzy integral provides a useful mechanism for evidence aggregation. It is a flexible... more The Choquet fuzzy integral provides a useful mechanism for evidence aggregation. It is a flexible method which can represent weighted averages, medians, order statistics, and many other information aggregation mechanisms. In this paper, two applications are described to handwritten word recognition: as a match function in a dynamic programming based classifier and as a method for fusing the results from multiple word recognition algorithms. In the first case, the results are compared with traditional methods. In the second case, the results are compared with neural network and Borda count approaches.
Radio Science, 2004
ABSTRACT The utility of acoustic-to-seismic coupling systems for landmine detection has been clea... more ABSTRACT The utility of acoustic-to-seismic coupling systems for landmine detection has been clearly established. In this approach, laser Doppler vibrometers (LDV) are used to measure the different responses to acoustic excitation in ground regions with and without buried landmines. Currently, for most applications, only the magnitude of the surface velocity is investigated and used to construct recognition algorithms. Recently, we introduced phase-based features in the classification scheme, significantly lowering false alarm rates at given detection probabilities. In this paper, we present modeling equations that explain the phase features for ground areas both with and without buried landmines from the perspective of harmonic oscillator models. We also describe the image processing techniques applied to velocity data collected in the time domain with a moving LDV array. The observed signatures are also compared with the prediction of the models described. We also construct classifiers with only magnitude information and both magnitude and phase information for this time-domain data set. Classification results indicate that we can combine magnitude and phase features to improve the detection of buried mines while reducing false alarms. We also find that using phase information improves the distinction between ground regions with buried landmines or man-made clutter objects.
Linear Algebra and its Applications, 1990
The Journal of the Acoustical Society of America, 2004
ABSTRACT Acoustic/seismic systems demonstrate significant potential for reliable detection of lan... more ABSTRACT Acoustic/seismic systems demonstrate significant potential for reliable detection of landmines, particularly when coupled with ground-penetrating radar (GPR). Acoustic/seismic systems provide complementary information to GPR. One difficulty in developing detection algorithms for acoustic/seismic systems is that the frequency bands at which information occurs vary according to factors such as mine depth and mechanical properties of the soil, which are not known during mine detection. Ordered weighted-averaging (OWA) systems offer the potential for robust detection of landmines under different conditions because they naturally provide the capability to process the best collection of frequency bands in a fashion that can be optimized. In this paper, a processing framework is presented that combines OWA operators for feature analysis with decision making. This framework can be optimized using acoustic/seismic signals as well as acoustic/seismic signals combined with GPR. In the latter case, the optimization can help to provide the appropriate mechanism for combining the complementary information from both sensors. Results are presented using real data from both mines and other buried objects collected at an outdoor test site.
Journal of Electronic Imaging, 2000
Journal of Electronic Imaging, 1996
ABSTRACT The Choquet fuzzy integral is applied to handwritten word recognition. A handwritten wor... more ABSTRACT The Choquet fuzzy integral is applied to handwritten word recognition. A handwritten word recognition system is described. The word recognition system assigns a recognition confidence value to each string in a lexicon of candidate strings. The system uses a lexicon-driven approach that integrates segmentation and recognition via dynamic programming matching. The dynamic programming matcher finds a segmentation of the word image for each string in the lexicon. The traditional match score between a segmentation and a string is an average. In this paper, fuzzy integrals are used instead of an average. Experimental results demonstrate the utility of this approach. A surprising result is obtained that indicates a simple choice of fuzzy integral works better than a more complex choice.
ABSTRACT The HSTAMIDS hand-held mine detection system can be operated in scan mode and investigat... more ABSTRACT The HSTAMIDS hand-held mine detection system can be operated in scan mode and investigation mode. In investigation mode, it is assumed that the system has produced an initial alarm at a location and the operator investigates the region around the location intensively. This mode of operation provides a different set of assumptions and requirements for detection algorithms operating within the unit. This paper discusses how the assumptions and requirements change and describes experimental laboratory results (using data collected in the field) in which significant improvements in Probability of Detection vs. False Alarm Rates were achieved. Two methods for region processing are discussed. Both methods transfer the frequency domain data acquired by the HSTAMIDS system into the time domain. One method is based on Size-Contrast Filtering and the other is based on robustly measuring the variation along the region corresponding to the potential mine location. These new methods are compared to the Correlation Detector, an online algorithm for detecting mines with the HSTAMIDS GPR which has been described in previous papers. Particular attention is given to deep Anti-Tank mines, which are difficult because they too far from the metal detector to produce a metal detector response. In this case, a factor of 6 reduction in FAR is obtained at 100% Probability of Detection.
ABSTRACT For explosive detection purposes, it is assumed that the per- son preparing or carrying ... more ABSTRACT For explosive detection purposes, it is assumed that the per- son preparing or carrying the explosive will inadvertently contaminate him/herself or the exterior of the package. To detect such traces of explosive materials, we show the use of differential reflectometry (DR) as an alternative system to the existing techniques. With DR, explosives show char- acteristic behaviours at specific wavelengths, for example, spectra of TNT shows a sudden decrease at 420 nm. To detect these behaviours, principle component analysis was performed to reduce the dimensionality of the data, and a support vector machine classifier was trained to identify TNT. With a 10-fold classification on 10000 non-TNT and 1935 TNT pixels, we achieved 0.3% false alarm rate at 75% true positive rate. In this study, we outline the operation of the DR system, show the unique signatures of explosives when viewed with DR, and report the detection rates with support vector machine classifiers.
SPIE Proceedings, 2007
We propose a general method for detecting landmine signatures in vehicle mounted ground penetrati... more We propose a general method for detecting landmine signatures in vehicle mounted ground penetrating radar (GPR) using discrete hidden Markov models and Gabor wavelet features. Observation vectors are constructed based on the expansion of the signature's B-scan using a bank of ...
Context-based unmixing has been studied by several re-searchers. Recent techniques, such as piece... more Context-based unmixing has been studied by several re-searchers. Recent techniques, such as piece-wise convex unmixing using fuzzy and possibilistic clustering or Bayesian methods proposed in [11] attempt to form contexts via clus-tering. It is assumed that the linear mixing model applies to each cluster (context) and endmembers and abundances are found for each cluster. As the clusters are spatially coher-ent, hyperspectral image segmentation can significantly aid unmixing approaches that perform cluster specific estimation of endmembers. In this work, we integrate a graph-cuts seg-mentation algorithm with piece-wise convex unmixing. This is compared to fuzzy clustering (FCM) with results obtained on two datasets. The results demonstrate that the integrated approach achieves better segmentation and more precise end-member identification (in terms of comparisons with known ground truth).
IEEE Geoscience and Remote Sensing Magazine, 2013
SPIE Proceedings, 2005
ABSTRACT Identifying unique patterns of energy in ground penetrating radar images plays an import... more ABSTRACT Identifying unique patterns of energy in ground penetrating radar images plays an important role in landmine/clutter discrimination. Many different geometric features, including size and the distribution of energy values in a radar image, can be exploited in mine detection and discrimination. The granulometry of a random set (image), computed by measuring the integral of a sequence of closings with elements from a family of increasing homothetic shapes, yields a size distribution that can be used for texture analysis or object detection and discrimination. An important complement to granulometries for discrimination is the compactometry, a feature we have identified that is computed by measuring the integral of a sequence of increasing concentric homothetic subsets of a random set. The compactometry yields a characterization of the concentration of the density distribution of the random set at a given point. This paper investigates the properties of compactometry and its derivative, the density spectrum, and demonstrates how they can be used together with granulometry to address the problem of landmine/clutter discrimination using ground-penetrating radar sensors.