Abstract: A system that passively scans for electromagnetic radiation across a potentially wide area and wide spectrum to identify objects in the area based on their emission spectra. The system may use antennas that can be aimed to sweep across a large range of azimuth and elevation angles, and frequency selectors that measure signals in a large set of frequency bands. To make scanning this large space feasible, the system uses recursive adaptive scanning, which scans at progressively finer grid spacings only in areas where signal intensities in each frequency band are changing rapidly. Preliminary scans may be made to measure background noise levels in each frequency and direction, again using recursive adaptive scanning, and this noise may be subtracted during scans for objects of interest. Objects may be identified by finding spikes in any of the frequency bands and matching them to a database of known object signatures.

Abstract: A hyperspectral facial analysis system and method for personalized health scoring to assess the risk that a person has a disease. Embodiments capture images in multiple spectral bands, such as visible, infrared, and ultraviolet, and analyze these images to generate multiple health metrics, such as pallor, temperature, sweat, and chromophores. These metrics may be combined into an overall health score that may be used for screening. Image analysis may focus on the area under the eyes, where skin is thinnest. Images may be compared to a reference population to identify anomalous values, so that health scoring automatically adjusts for local conditions. Pallor may be calculated based on hue and saturation of visible light images. Temperature may be calculated based on infrared image intensity. Sweat may be calculated using cross-polarized images to identify specular highlights. Chromophores may be calculated by comparing frequency domain ultraviolet images to those of the reference population.

Abstract: A hyperspectral facial analysis system and method for personalized health scoring to assess the risk that a person has a disease. Embodiments capture images in multiple spectral bands, such as visible, infrared, and ultraviolet, and analyze these images to generate multiple health metrics, such as pallor, temperature, sweat, and chromophores. These metrics may be combined into an overall health score that may be used for screening. Image analysis may focus on the area under the eyes, where skin is thinnest. Images may be compared to a reference population to identify anomalous values, so that health scoring automatically adjusts for local conditions. Pallor may be calculated based on hue and saturation of visible light images. Temperature may be calculated based on infrared image intensity. Sweat may be calculated using cross-polarized images to identify specular highlights. Chromophores may be calculated by comparing frequency domain ultraviolet images to those of the reference population.

Abstract: Hand recognition system that compares narrow band ultraviolet-absorbing skin chromophores to identify a subject. Ultraviolet images of hands show much greater detail than visible light images, so matching of ultraviolet images may be much more accurate. A database of known persons may contain reference ultraviolet hand images tagged with each person's identity. Reference images and subject images may be processed to locate the hands, identify features (such as chromophores), compare and match feature descriptors, and calculate correlation scores between the subject image and each reference image. Locating and normalizing hand images may use infrared and visible light cameras in addition to ultraviolet. If the subject is moving, the subject's hand may be tracked, a 3D model of the subject's hand may be developed from multiple images, and this model may be rotated so that the orientation matches that of the reference images.

Abstract: Facial recognition system that compares narrow band ultraviolet-absorbing skin chromophores to identify a subject. Ultraviolet images show much greater facial detail than visible light images, so matching of ultraviolet images may be much more accurate. Embodiments of the system may have a camera that is sensitive to ultraviolet, and a special lens and filter that pass the relevant ultraviolet wavelengths. A database of known persons may contain reference ultraviolet facial images tagged with each person's identity. Reference images and subject images may be processed to locate the face, identify features (such as chromophores), compare and match feature descriptors, and calculate correlation scores between the subject image and each reference image. If the subject is moving, the subject's face may be tracked, a 3D model of the subject's face may be developed from multiple images, and this model may be rotated so that the orientation matches that of the reference images.

Abstract: A hyperspectral facial analysis system and method for personalized health scoring to assess the risk that a person has a disease. Embodiments capture images in multiple spectral bands, such as visible, infrared, and ultraviolet, and analyze these images to generate multiple health metrics, such as pallor, temperature, sweat, and chromophores. These metrics may be combined into an overall health score that may be used for screening. Image analysis may focus on the area under the eyes, where skin is thinnest. Images may be compared to a reference population to identify anomalous values, so that health scoring automatically adjusts for local conditions. Pallor may be calculated based on hue and saturation of visible light images. Temperature may be calculated based on infrared image intensity. Sweat may be calculated using cross-polarized images to identify specular highlights. Chromophores may be calculated by comparing frequency domain ultraviolet images to those of the reference population.

Abstract: System that transforms 2D images of slices of an object, such as a tissue sample from a biopsy, into a 3D point cloud for visualization. 3D points may be generated for pixels in the 2D images with high luminance. Point depths may be assigned pseudo-randomly so that the points fill the space within the slice. The probability distribution for the random depths may be based on any characteristics of the 2D images. For example, points corresponding to pixels within large areas of high luminance may be spread relatively uniformly through the slice, while isolated points or points in small areas may be biased towards the minimum depth of the slice. Points in the point cloud may be partitioned into channels, corresponding for example to different stains of a sample, and may be visualized for example as different color assigned to points.