Abstract: Within examples, methods and systems for occlusion-robust object fingerprinting using fusion of multiple sub-region signatures are described. An example method includes receiving an indication of an object within a sequence of video frames, selecting from the sequence of video frames a reference image frame indicative of the object and candidate image frames representative of possible portions of the object, dividing the reference image frame and the candidate image frames into multiple cells, defining for the reference image frame and the candidate image frames sub-regions of the multiple cells such that the sub-regions include the same cells for overlapping representations and the sub-regions include multiple sizes, comparing characteristics of sub-regions of the reference image frame to characteristics of sub-regions of the candidate image frames and determining similarity measurements, and based on the similarity measurements, tracking the object within the sequence of video frames.

Abstract: The present disclosure generally relates to systems and methods for spectral demixing. An example technique includes obtaining empirical spectroscopic data representing a plurality of frequencies of electromagnetic energy that has interacted with a specimen, accessing a computer readable representation of a hierarchal spectral cluster tree representing a spectral library, demixing, with data on each of a plurality of levels of the hierarchal spectral cluster tree, foveated spectroscopic data derived from the empirical spectroscopic data, identifying at least one node in the hierarchal spectral cluster tree as corresponding to the empirical spectroscopic data, and outputting an endmember abundance assessment of the specimen corresponding to at least the at least one node.

Abstract: The present invention relates to a classifier cascade object detection system. The system operates by inputting an image patch into parallel feature generation modules, each of the feature generation modules operable for extracting features from the image patch. The features are provided to an opportunistic classifier cascade, the opportunistic classifier cascade having a series of classifier stages. The opportunistic classifier cascade is executed by progressively evaluating, in each classifier in the classifier cascade, the features to produce a response, with each response progressively utilized by a decision function to generate a stage response for each classifier stage. If each stage response exceeds a stage threshold then the image patch is classified as a target object, and if the stage response from any of the decision functions does not exceed the stage threshold, then the image patch is classified as a non-target object.

Abstract: Described herein is a method for enhancing image data that includes transforming image data from an intensity domain to a wavelet domain to produce wavelet coefficients. A first set of wavelet coefficients of the wavelet coefficients includes low-frequency wavelet coefficients. The method also includes modifying the first set of wavelet coefficients using a coefficient distribution based filter to produce a modified first set of wavelet coefficients. The method includes transforming the modified first set of wavelet coefficients from the wavelet domain to the intensity domain to produce enhanced image data.

Abstract: A method includes receiving first tracking data and a first confidence value at a tracker selection system from a first tracking system. The method includes receiving second tracking data and a second confidence value at the tracker selection system from a second tracking system. The tracking systems may be configured to track an object in a sequence of images, and the first and second tracking data may indicate locations of regions where the corresponding tracking system has tracked the object in an image of the sequence of images. The confidence values may indicate likelihoods that a corresponding tracking system is tracking the object. The method further includes providing output data to the first tracking system and to the second tracking system. The output data may include data selected from the first tracking data and the second tracking data based on a result of a comparison of the confidence values.

Abstract: A method includes receiving image data at a first tracking system. The image data may represent a region in an image of a sequence of images. The method includes generating a first tracking fingerprint based on the image data. The method includes comparing the first tracking fingerprint and a second tracking fingerprint. The method further includes providing an output from the first tracking system to a second tracking system based on a result of the comparison of the first tracking fingerprint and the second tracking fingerprint.

Abstract: Described is a high-dimensional optimization system implementing a modification of particle swarm optimization called self-aware particle swarm optimization. A plurality of software agents is configured to operate as a cooperative swarm to locate an objective function optima in a multi-dimensional solution space. The plurality of software agents is influenced by a set of parameters which influence exploration of the multi-dimensional solution space and convergence on the objective function optima. The plurality of software agents automatically modifies the set of parameters in response to at least one measure of convergence progress. Self-aware particle swarm optimization allows for monitoring of simple convergence properties to provide feedback to the swarm dynamics and make the swarm self-aware and adjust itself to the problem being solved.

Abstract: Systems and methods are provided for backfilling a 3D cloud of coordinates. One embodiment is an apparatus that includes a camera able to capture an image, and a ranging system able to measure distances objects. The apparatus also includes a controller able to analyze the measured distances to generate a cloud of three dimensional coordinates, and to project coordinates of the cloud onto the image to match the coordinates with pixels of the image. The controller is also able to generate additional coordinates to increase the resolution of the cloud by iteratively defining a scanning window, detecting a group of coordinates that have been projected onto the scanning window, scoring each coordinate in the group based on its similarity to other coordinates in the group, dividing the scanning window into quadrants, selecting a coordinate from each quadrant, and generating an additional coordinate based on the selected coordinates.

Abstract: A method for detecting an opening in a structure represented by a three-dimensional point cloud may include the steps of: (1) creating a three-dimensional point cloud map of a scene, the three-dimensional point cloud map including a plurality of points representing a ground plane and the structure upon the ground plane, (2) identifying an absence of points within the plurality of points representing the structure, and (3) determining whether the absence of points represents the opening in the structure.

Abstract: Described is a system for adaptive three-dimensional (3D) to two-dimensional (2D) projection for different height slices and extraction of morphological features. Initially, the system receives 3D point cloud data. Next, an image pixel number to point cloud number ratio is selected. Thereafter, an image row and image column are selected to identify desired height slices. The 3D point cloud is then accumulated on the desired height slices to generate a plurality of 2D height slices.

Abstract: A method for detecting an opening in a structure represented by a three-dimensional point cloud may include the steps of: (1) creating a three-dimensional point cloud map of a scene, the three-dimensional point cloud map including a plurality of points representing a ground plane and the structure upon the ground plane, (2) identifying an absence of points within the plurality of points representing the structure, and (3) determining whether the absence of points represents the opening in the structure.

Abstract: Described is system and method for blind beamforming using knowledge embedded in transmitted signals. Initial antenna weights are assigned for antenna elements of emitters of a beamforming system to generate a set of weighted radio-frequency (RF) signals at an output of each antenna element. The set of weighted RF signals are combined to form RF signal mixtures. Then, each RF signal mixture is processed to extract embedded information within signals of the emitters. The extracted embedded information is sent to an optimization module, where the extracted embedded information is used to perform simultaneous signal extractions for emitters in the beamforming system. Furthermore, feedback from the optimization module is used to modify antenna weights toward optimal beamforming.

Abstract: A human monitoring system includes a plurality of cameras and a visual processor. The plurality of cameras are disposed about a workspace area, where each camera is configured to capture a video feed that includes a plurality of image frames, and the plurality of image frames are time-synchronized between the respective cameras. The visual processor is configured to receive the plurality of image frames from the plurality of vision-based imaging devices and determine an integrity score for each respective image frame. The processor may then isolate a foreground section from two or more of the views, determine a principle body axis for each respective foreground section, and determine a location point according to a weighted least squares function amongst the various principle body axes.

Abstract: A method and target detector for detecting targets. A number of bright pixels are identified in an image. Each of the number of bright pixels belongs to a line detected in the image in which the line represents a candidate for a target. A number of feature vectors are identified for the number of bright pixels in the image. Each of the number of feature vectors is classified as one of a target vector representing the target and a clutter vector representing clutter.

Abstract: Described is a system for foveated compressive sensing. The system is configured to receive an input image f of a scene and initialize a measurement matrix. Global measurements are then performed, with a lower resolution image of the scene thereafter reconstructed. Task salient regions are extracted from the low resolution image. Thereafter, the system estimates a task-specific operator and detects regions-of-interest (ROI) based on the task salient regions. An ROI-adapted and foveated measurement matrix is then generated. Local measurements are then performed on task-relevant ROIs. A higher resolution image can then be reconstructed of the scene to allow for identification of objects in the ROI.

Abstract: A method for increasing resolution of an image formed of received light from an illuminated spot includes measuring a y vector for measurement kernels A1to AM, where M is a number of the measurement kernels, measuring the y vector including programming a programmable N-pixel micromirror or mask located in a return path of a received reflected scene spot with a jth measurement kernel Aj of the measurement kernels A1 to AM, measuring y, wherein y is an inner product of a scene reflectivity f(?,?) with the measurement kernel Aj for each range bin ri, wherein ? and ? are azimuth and elevation angles, respectively, repeating programming the programmable N-pixel micromirror or mask and measuring y for each measurement kernel A1 to AM, and forming a reconstructed image using the measured y vector, wherein forming the reconstructed image includes using compressive sensing or Moore-Penrose reconstruction.

Abstract: A method for detecting one or more target objects is provided including obtaining 2-dimensional imaging information and 3-dimensional point cloud information of a target zone. The method also includes determining a ground plane in the point cloud information and removing the ground plane to generate modified 3-dimensional information. Also, the method includes identifying a set of 2-dimensional candidate objects from the 2-dimensional imaging information, and identifying a set of 3-dimensional candidate objects from the modified 3-dimensional information. The method also includes determining, for each of at least some of the 2-dimensional candidate objects, a corresponding 3-dimensional candidate object from the set of 3-dimensional candidate objects.

Abstract: A method for detecting one or more target objects is provided including obtaining 2-dimensional imaging information and 3-dimensional point cloud information of a target zone. The method also includes determining a ground plane in the point cloud information and removing the ground plane to generate modified 3-dimensional information. Also, the method includes identifying a set of 2-dimensional candidate objects from the 2-dimensional imaging information, and identifying a set of 3-dimensional candidate objects from the modified 3-dimensional information. The method also includes determining, for each of at least some of the 2-dimensional candidate objects, a corresponding 3-dimensional candidate object from the set of 3-dimensional candidate objects.

Abstract: A method of constructing a probabilistic representation of the location of an object within a workspace includes obtaining a plurality of 2D images of the workspace, with each respective 2D image being acquired from a camera disposed at a different location within the workspace. A foreground portion is identified within at least two of the plurality of 2D images, and each foreground portion is projected to each of a plurality of parallel spaced planes. An area is identified within each of the plurality of planes where a plurality of projected foreground portions overlap. These identified areas are combined to form a 3D bounding envelope of an object. This bounding envelope is a probabilistic representation of the location of the object within the workspace.

Abstract: A human monitoring system includes a plurality of cameras and a visual processor. The plurality of cameras are disposed about a workspace area, where each camera is configured to capture a video feed that includes a plurality of image frames, and the plurality of image frames are time-synchronized between the respective cameras. The visual processor is configured to identify the presence of a human within the workspace area from the plurality of image frames, generate a motion track of the human within the workspace area, generate an activity log of one or more activities performed by the human throughout the motion track, and compare the motion track and activity log to an activity template that defines a plurality of required actions. The processor then provides an alert if one or more actions within the activity template are not performed within the workspace area.