Allen Waxman - Academia.edu (original) (raw)

Papers by Allen Waxman

Research paper thumbnail of A neural system for behavioral conditioning of mobile robots

1990 IJCNN International Joint Conference on Neural Networks

ABSTRACT

Research paper thumbnail of Parametric Study of Diffusion-Enhancement Networks for Spatiotemporal Grouping in Real-Time Artificial Vision

Lincoaln ;aboratory MASSACHUSETTS INSTITUTE OF TECHNOLOGY LEXINCTON, 41ASNACHUSJEVS Prepared f6'r... more Lincoaln ;aboratory MASSACHUSETTS INSTITUTE OF TECHNOLOGY LEXINCTON, 41ASNACHUSJEVS Prepared f6'r the Department of the Air Force under Contract F]962&-90-C4K)ft2, Approived for pubik reirase; dio~ributhin in unlimited. MTIS CRA&t A -I t[;lC TA8 El Approved for public release; distribution is unlimited. By Distribution I Availability Codes Avail dridlor [its? soecial LEXINGTON MASSACHUSETTS

Research paper thumbnail of Multisensor & spectral image fusion & mining: from neural systems to applications

32nd Applied Imagery Pattern Recognition Workshop, 2003. Proceedings., 2000

We have continued development of a system for multisensor image fusion and interactive mining bas... more We have continued development of a system for multisensor image fusion and interactive mining based on neural models of color vision processing, learning and pattern recognition. We pioneered this work while at MIT Lincoln Laboratory, initially for color fused night vision (low-light visible and uncooled thermal imagery) and later extended it to multispectral IR and 3D ladar. We also developed a proof-of-concept system for EO, IR, SAR fusion and mining. Over the last year we have generalized this approach and developed a user-friendly system integrated into a COTS exploitation environment known as ERDAS Imagine. In this paper, we will summarize the approach and the neural networks used, and demonstrate fusion and interactive mining (i.e., target learning and search) of low-light visible/SWIR/MWIR/LWIR night imagery, and IKONOS multispectral and high-resolution panchromatic imagery. In addition, we will demonstrate how target learning and search can be enabled over extended operating conditions by allowing training over multiple scenes. This will be illustrated for the detection of small boats in coastal waters using fused visible/MWIR/LWIR imagery.

Research paper thumbnail of Electronic imaging aids for night driving: low-light CCD, thermal IR, and color fused visible/IR

Storage and Retrieval For Image and Video Databases, 1996

ABSTRACT MIT Lincoln Laboratory is developing new electronic night vision technologies for defens... more ABSTRACT MIT Lincoln Laboratory is developing new electronic night vision technologies for defense applications which can be adapted for civilian applications such as night driving aids. These technologies include (1) low-light CCD imagers capable of operating under starlight illumination conditions at video rates, (2) realtime processing of wide dynamic range imagery (visible and IR) to enhance contrast and adaptively compress dynamic range, and (3) realtime fusion of low-light visible and thermal IR imagery to provide color display of the night scene to the operator in order to enhance situational awareness. This paper compares imagery collected during night driving including: low-light CCD visible imagery, intensified-CCD visible imagery, uncooled long-wave IR imagery, cryogenically cooled mid-wave IR imagery, and visible/IR dual-band imagery fused for gray and color display.

Research paper thumbnail of An Image Flow Paradigm

Research paper thumbnail of 1994 Special Issue: Mobile robot visual mapping and localization: A view-based neurocomputational architecture that emulates hippocampal place learning

Neural Networks, Sep 1, 1994

Research paper thumbnail of Combining evidence from multiple views of 3-D objects

Proc Spie, Apr 1, 1992

ABSTRACT

Research paper thumbnail of Fusion of visible and thermal imagery improves situational awareness. Discussion. Author's reply

Agard Conference Proceedings, 1998

Résumé/Abstract A new colour image fusion scheme is applied to visible and thermal images of mili... more Résumé/Abstract A new colour image fusion scheme is applied to visible and thermal images of military relevant scenarios. An observer experiment is performed to test if the increased amount of detail in the fused images can improve the accuracy of observers performing a detection and localisation task. The results show that observers can localise a target in a scene (1) with a significantly higher accuracy, and (2) with a greater amount of confidence when they perform with fused images (either gray or colour fused), compared ...

Research paper thumbnail of A prototype system for 3D color fusion and mining of multisensor/spectral imagery

We have developed a prototype system in which a user can fuse up to 4 modalities (or 4 spectral b... more We have developed a prototype system in which a user can fuse up to 4 modalities (or 4 spectral bands) of imagery previously registered to one another with respect to a 3D terrain model. The color fused imagery can be draped onto the terrain to support interactive 3D flythrough. The fused imagery, and its opponent-sensor contrasts, can be further processed to yield extended boundary contours and texture measures. Together, these layers of registered imagery and image features can be interactively mined for objects of interest. Data mining for infrastructure and compact targets is achieved using a point-and-click user interface in conjunction with a Fuzzy ARTMAP neural network for on-line pattern learning and recognition. Graphical user interfaces enable the user to control each stage of processing: image enhancement, image fusion, contour and texture extraction, 3D terrain characterization, 3D graphics model building, preparation for exploitation, and interactive data mining. The system is configured as a client-server architecture, enabling remote collaborative exploitation of multisensor imagery. Throughout, the processing of imagery and patterns relies on neural network models of spatial and color opponency, and the adaptive resonance theory of pattern processing. This system has been used to process imagery of a variety of geographic sites, in order to extract roads, rivers, forests and orchards, and performance has been assessed against manually determined "ground truth." The data mining approach has been extended for the case of hyperspectral imagery of hundreds of bands. This prototype system has now been installed at multiple US government sites for evaluation by image analysts. We plan to extend this approach to include various nonimaging sensor modalities that can be localized to geographic coordinates (e.g., GMTI and SIGINT). We also plan to embed these image fusion and mining capabilities in commercial open software environments for image processing and GIS.

Research paper thumbnail of Automated activity pattern learning and monitoring provide decision support to supervisors of busy environments

Intelligent Decision Technologies, 2009

Neurobiologically inspired algorithms for exploiting track data to learn normal patterns of motio... more Neurobiologically inspired algorithms for exploiting track data to learn normal patterns of motion behavior, detect deviations from normalcy, and predict future behavior are presented. These capabilities contribute to higher-level fusion situational awareness and assessment ...

Research paper thumbnail of Modular Processing Stages of the Pipe Machine

Research paper thumbnail of Real time adaptive digital image processing for dynamic range remapping of imagery including low-light-level visible imagery

Research paper thumbnail of Image fusion & mining tools for a COTS environment

Sixth International Conference of Information Fusion, 2003. Proceedings of the, 2003

We have continued development of a system for multisensor image fusion and interactive mining bas... more We have continued development of a system for multisensor image fusion and interactive mining based on neural models of color vision processing, learning and pattern recognition. We pioneered this work while at MIT Lincoln Laboratory, initially for color fused night vision (low-light visible and uncooled thermal imagery) and later extended it to multispectral IR and 3D ladar. We also developed a proof-of-concept system for EO, IR, SAR fusion and mining. Over the last year we have generalized this approach and developed a userfriendly system integrated into a COTS exploitation environment known as ERDAS Imagine. In this paper, we will summarize the approach and the neural networks used, and demonstrate fusion and interactive mining (i.e., target learning and search) of low-light visible/SWIR/MWIR/LWIR night imagery, and IKONOS multispectral and high-resolution panchromatic imagery. In addition, we will demonstrate how target learning and search can be enabled over extended operating conditions by allowing training over multiple scenes. This will be illustrated for the detection of small boats in coastal waters using fused visible/MWIR/LWIR imagery.

Research paper thumbnail of The analytic structure of image flows: Deformation and segmentation

Computer Vision, Graphics, and Image Processing, 1990

Time-varying imagery is often described in terms of image flow fields (i.e., image motion), which... more Time-varying imagery is often described in terms of image flow fields (i.e., image motion), which correspond to the perceptive projection of feature motions in three dimensions (3D). In the case of multiple moving objects with smooth surfaces, the image flow possesses an analytic structure that reflects these 3D properties. This paper describes the analytic structure of image flow fields in the image space-time domain, and its use for segmentation and 3D motion computation. First we discuss the localflow structure as embodied in the concept of neighborhood deformation. The local image deformation is effectively represented by a set of 12 basis deformations, each of which is responsible for an independent deformation. This local representation provides us with sufficient information for the recovery of 3D object structure and motion, in the case of relative rigid body motions. We next discuss the globalflow structure embodied in the partitioning of the entire image plane into analytic regions separated by boundaries of analyticity, such that each small neighborhood within the analytic region is described in terms of deformation bases. This analysis reveals an effective mechanism for detecting the analytic boundaries of flow fields, thereby segmenting the image into meaningful regions. The notion of consistency which is often used in the image segmentation is made explicit by the mathematical notion of analyticity derived from the projection relation of 3D object motion. The concept of flow analyticity is then extended to the temporal domain, suggesting a more robust algorithm for recovering image flow from multiple frames. Finally, we argue that the process of flow segmentation can be understood in the framework of grouping process. The general concept of coherence or grouping through local support (such as the second-order flows in our case) is discussed.

Research paper thumbnail of Dynamics of a Couple-Stress Fluid Membrane

Studies in Applied Mathematics, 1984

ABSTRACT

Research paper thumbnail of Closed form solutions to image flow equations for planar surfaces in motion

Computer Vision Graphics and Image Processing, Nov 1, 1986

ABSTRACT

Research paper thumbnail of Neural Methods for Imagery, GMTI, and Information Fusion

... synchronization of node activity, which generic urban environment has been constructed which ... more ... synchronization of node activity, which generic urban environment has been constructed which can be viewed as temporal binding of these nodes in short-includes HyMap hyperspectral imagery, high-resolution term memory (STM). Pair-wise associative learning EO imagery ...

Research paper thumbnail of Opponent-Color Fusion of Multi-Sensor Imagery: Visible, IR and SAR

... displays (Widdel & Pfendler, 1990). It is common experience (eg, watching television and ... more ... displays (Widdel & Pfendler, 1990). It is common experience (eg, watching television and movies) that color imagery generates greater and more rapid scene comprehension than grayscale imagery does. We have every reason to expect that color fused imagery of high quality ...

Research paper thumbnail of Information fusion for image analysis: geospatial foundations for higher-level fusion

Proceedings of the Fifth International Conference on Information Fusion. FUSION 2002. (IEEE Cat.No.02EX5997), 2000

In support of the AFOSR program in Information Fusion, the CNS Technology Laboratory at Boston Un... more In support of the AFOSR program in Information Fusion, the CNS Technology Laboratory at Boston University is developing and applying neural models of image and signal processing, pattern learning and recognition, associative learning dynamics, and 3D visualization, to the domain of Information Fusion for Image Analysis in a geospatial context. Our research is focused by a challenge problem involving the emergence of a crisis in an urban environment, brought on by a terrorist attack or other man-made or natural disaster. We aim to develop methods aiding preparation and monitoring of the battlespace, deriving context from multiple sources of imagery (high-resolution visible and low-resolution hyperspectral) and signals (GMTI from moving vehicles, and ELINT from emitters). This context will serve as a foundation, in conjunction with existing knowledge nets, for exploring neural methods in higherlevel information fusion supporting situation assessment and creation of a common operating picture (COP).

Research paper thumbnail of An Image Flow Simulator

Research paper thumbnail of A neural system for behavioral conditioning of mobile robots

1990 IJCNN International Joint Conference on Neural Networks

ABSTRACT

Research paper thumbnail of Parametric Study of Diffusion-Enhancement Networks for Spatiotemporal Grouping in Real-Time Artificial Vision

Lincoaln ;aboratory MASSACHUSETTS INSTITUTE OF TECHNOLOGY LEXINCTON, 41ASNACHUSJEVS Prepared f6'r... more Lincoaln ;aboratory MASSACHUSETTS INSTITUTE OF TECHNOLOGY LEXINCTON, 41ASNACHUSJEVS Prepared f6'r the Department of the Air Force under Contract F]962&-90-C4K)ft2, Approived for pubik reirase; dio~ributhin in unlimited. MTIS CRA&t A -I t[;lC TA8 El Approved for public release; distribution is unlimited. By Distribution I Availability Codes Avail dridlor [its? soecial LEXINGTON MASSACHUSETTS

Research paper thumbnail of Multisensor & spectral image fusion & mining: from neural systems to applications

32nd Applied Imagery Pattern Recognition Workshop, 2003. Proceedings., 2000

We have continued development of a system for multisensor image fusion and interactive mining bas... more We have continued development of a system for multisensor image fusion and interactive mining based on neural models of color vision processing, learning and pattern recognition. We pioneered this work while at MIT Lincoln Laboratory, initially for color fused night vision (low-light visible and uncooled thermal imagery) and later extended it to multispectral IR and 3D ladar. We also developed a proof-of-concept system for EO, IR, SAR fusion and mining. Over the last year we have generalized this approach and developed a user-friendly system integrated into a COTS exploitation environment known as ERDAS Imagine. In this paper, we will summarize the approach and the neural networks used, and demonstrate fusion and interactive mining (i.e., target learning and search) of low-light visible/SWIR/MWIR/LWIR night imagery, and IKONOS multispectral and high-resolution panchromatic imagery. In addition, we will demonstrate how target learning and search can be enabled over extended operating conditions by allowing training over multiple scenes. This will be illustrated for the detection of small boats in coastal waters using fused visible/MWIR/LWIR imagery.

Research paper thumbnail of Electronic imaging aids for night driving: low-light CCD, thermal IR, and color fused visible/IR

Storage and Retrieval For Image and Video Databases, 1996

ABSTRACT MIT Lincoln Laboratory is developing new electronic night vision technologies for defens... more ABSTRACT MIT Lincoln Laboratory is developing new electronic night vision technologies for defense applications which can be adapted for civilian applications such as night driving aids. These technologies include (1) low-light CCD imagers capable of operating under starlight illumination conditions at video rates, (2) realtime processing of wide dynamic range imagery (visible and IR) to enhance contrast and adaptively compress dynamic range, and (3) realtime fusion of low-light visible and thermal IR imagery to provide color display of the night scene to the operator in order to enhance situational awareness. This paper compares imagery collected during night driving including: low-light CCD visible imagery, intensified-CCD visible imagery, uncooled long-wave IR imagery, cryogenically cooled mid-wave IR imagery, and visible/IR dual-band imagery fused for gray and color display.

Research paper thumbnail of An Image Flow Paradigm

Research paper thumbnail of 1994 Special Issue: Mobile robot visual mapping and localization: A view-based neurocomputational architecture that emulates hippocampal place learning

Neural Networks, Sep 1, 1994

Research paper thumbnail of Combining evidence from multiple views of 3-D objects

Proc Spie, Apr 1, 1992

ABSTRACT

Research paper thumbnail of Fusion of visible and thermal imagery improves situational awareness. Discussion. Author's reply

Agard Conference Proceedings, 1998

Résumé/Abstract A new colour image fusion scheme is applied to visible and thermal images of mili... more Résumé/Abstract A new colour image fusion scheme is applied to visible and thermal images of military relevant scenarios. An observer experiment is performed to test if the increased amount of detail in the fused images can improve the accuracy of observers performing a detection and localisation task. The results show that observers can localise a target in a scene (1) with a significantly higher accuracy, and (2) with a greater amount of confidence when they perform with fused images (either gray or colour fused), compared ...

Research paper thumbnail of A prototype system for 3D color fusion and mining of multisensor/spectral imagery

We have developed a prototype system in which a user can fuse up to 4 modalities (or 4 spectral b... more We have developed a prototype system in which a user can fuse up to 4 modalities (or 4 spectral bands) of imagery previously registered to one another with respect to a 3D terrain model. The color fused imagery can be draped onto the terrain to support interactive 3D flythrough. The fused imagery, and its opponent-sensor contrasts, can be further processed to yield extended boundary contours and texture measures. Together, these layers of registered imagery and image features can be interactively mined for objects of interest. Data mining for infrastructure and compact targets is achieved using a point-and-click user interface in conjunction with a Fuzzy ARTMAP neural network for on-line pattern learning and recognition. Graphical user interfaces enable the user to control each stage of processing: image enhancement, image fusion, contour and texture extraction, 3D terrain characterization, 3D graphics model building, preparation for exploitation, and interactive data mining. The system is configured as a client-server architecture, enabling remote collaborative exploitation of multisensor imagery. Throughout, the processing of imagery and patterns relies on neural network models of spatial and color opponency, and the adaptive resonance theory of pattern processing. This system has been used to process imagery of a variety of geographic sites, in order to extract roads, rivers, forests and orchards, and performance has been assessed against manually determined "ground truth." The data mining approach has been extended for the case of hyperspectral imagery of hundreds of bands. This prototype system has now been installed at multiple US government sites for evaluation by image analysts. We plan to extend this approach to include various nonimaging sensor modalities that can be localized to geographic coordinates (e.g., GMTI and SIGINT). We also plan to embed these image fusion and mining capabilities in commercial open software environments for image processing and GIS.

Research paper thumbnail of Automated activity pattern learning and monitoring provide decision support to supervisors of busy environments

Intelligent Decision Technologies, 2009

Neurobiologically inspired algorithms for exploiting track data to learn normal patterns of motio... more Neurobiologically inspired algorithms for exploiting track data to learn normal patterns of motion behavior, detect deviations from normalcy, and predict future behavior are presented. These capabilities contribute to higher-level fusion situational awareness and assessment ...

Research paper thumbnail of Modular Processing Stages of the Pipe Machine

Research paper thumbnail of Real time adaptive digital image processing for dynamic range remapping of imagery including low-light-level visible imagery

Research paper thumbnail of Image fusion & mining tools for a COTS environment

Sixth International Conference of Information Fusion, 2003. Proceedings of the, 2003

We have continued development of a system for multisensor image fusion and interactive mining bas... more We have continued development of a system for multisensor image fusion and interactive mining based on neural models of color vision processing, learning and pattern recognition. We pioneered this work while at MIT Lincoln Laboratory, initially for color fused night vision (low-light visible and uncooled thermal imagery) and later extended it to multispectral IR and 3D ladar. We also developed a proof-of-concept system for EO, IR, SAR fusion and mining. Over the last year we have generalized this approach and developed a userfriendly system integrated into a COTS exploitation environment known as ERDAS Imagine. In this paper, we will summarize the approach and the neural networks used, and demonstrate fusion and interactive mining (i.e., target learning and search) of low-light visible/SWIR/MWIR/LWIR night imagery, and IKONOS multispectral and high-resolution panchromatic imagery. In addition, we will demonstrate how target learning and search can be enabled over extended operating conditions by allowing training over multiple scenes. This will be illustrated for the detection of small boats in coastal waters using fused visible/MWIR/LWIR imagery.

Research paper thumbnail of The analytic structure of image flows: Deformation and segmentation

Computer Vision, Graphics, and Image Processing, 1990

Time-varying imagery is often described in terms of image flow fields (i.e., image motion), which... more Time-varying imagery is often described in terms of image flow fields (i.e., image motion), which correspond to the perceptive projection of feature motions in three dimensions (3D). In the case of multiple moving objects with smooth surfaces, the image flow possesses an analytic structure that reflects these 3D properties. This paper describes the analytic structure of image flow fields in the image space-time domain, and its use for segmentation and 3D motion computation. First we discuss the localflow structure as embodied in the concept of neighborhood deformation. The local image deformation is effectively represented by a set of 12 basis deformations, each of which is responsible for an independent deformation. This local representation provides us with sufficient information for the recovery of 3D object structure and motion, in the case of relative rigid body motions. We next discuss the globalflow structure embodied in the partitioning of the entire image plane into analytic regions separated by boundaries of analyticity, such that each small neighborhood within the analytic region is described in terms of deformation bases. This analysis reveals an effective mechanism for detecting the analytic boundaries of flow fields, thereby segmenting the image into meaningful regions. The notion of consistency which is often used in the image segmentation is made explicit by the mathematical notion of analyticity derived from the projection relation of 3D object motion. The concept of flow analyticity is then extended to the temporal domain, suggesting a more robust algorithm for recovering image flow from multiple frames. Finally, we argue that the process of flow segmentation can be understood in the framework of grouping process. The general concept of coherence or grouping through local support (such as the second-order flows in our case) is discussed.

Research paper thumbnail of Dynamics of a Couple-Stress Fluid Membrane

Studies in Applied Mathematics, 1984

ABSTRACT

Research paper thumbnail of Closed form solutions to image flow equations for planar surfaces in motion

Computer Vision Graphics and Image Processing, Nov 1, 1986

ABSTRACT

Research paper thumbnail of Neural Methods for Imagery, GMTI, and Information Fusion

... synchronization of node activity, which generic urban environment has been constructed which ... more ... synchronization of node activity, which generic urban environment has been constructed which can be viewed as temporal binding of these nodes in short-includes HyMap hyperspectral imagery, high-resolution term memory (STM). Pair-wise associative learning EO imagery ...

Research paper thumbnail of Opponent-Color Fusion of Multi-Sensor Imagery: Visible, IR and SAR

... displays (Widdel & Pfendler, 1990). It is common experience (eg, watching television and ... more ... displays (Widdel & Pfendler, 1990). It is common experience (eg, watching television and movies) that color imagery generates greater and more rapid scene comprehension than grayscale imagery does. We have every reason to expect that color fused imagery of high quality ...

Research paper thumbnail of Information fusion for image analysis: geospatial foundations for higher-level fusion

Proceedings of the Fifth International Conference on Information Fusion. FUSION 2002. (IEEE Cat.No.02EX5997), 2000

In support of the AFOSR program in Information Fusion, the CNS Technology Laboratory at Boston Un... more In support of the AFOSR program in Information Fusion, the CNS Technology Laboratory at Boston University is developing and applying neural models of image and signal processing, pattern learning and recognition, associative learning dynamics, and 3D visualization, to the domain of Information Fusion for Image Analysis in a geospatial context. Our research is focused by a challenge problem involving the emergence of a crisis in an urban environment, brought on by a terrorist attack or other man-made or natural disaster. We aim to develop methods aiding preparation and monitoring of the battlespace, deriving context from multiple sources of imagery (high-resolution visible and low-resolution hyperspectral) and signals (GMTI from moving vehicles, and ELINT from emitters). This context will serve as a foundation, in conjunction with existing knowledge nets, for exploring neural methods in higherlevel information fusion supporting situation assessment and creation of a common operating picture (COP).

Research paper thumbnail of An Image Flow Simulator