Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Pixel”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 793 records · Page 44
All sourcesNASA NTRSDOE OSTI
NASA NTRS2012

Resolution Enhancement of MODIS-Derived Water Indices for Studying Persistent Flooding

Monitoring coastal marshes for persistent flooding and salinity stress is a high priority issue in Louisiana. Remote sensing can identify environmental variables that can be indicators of marsh habitat conditions, and offer timely and relatively accurate information for aiding wetland vegetation management. Monitoring activity accuracy is often limited by mixed pixels which occur when areas represented by the pixel encompasses more than one cover type. Mixtures of marsh grasses and open water in 250m Moderate Resolution Imaging Spectroradiometer (MODIS) data can impede flood area estimation. Flood mapping of such mixtures requires finer spatial resolution data to better represent the cover type composition within 250m MODIS pixel. Fusion of MODIS and Landsat can improve both spectral and temporal resolution of time series products to resolve rapid changes from forcing mechanisms like hurricane winds and storm surge. For this study, using a method for estimating sub-pixel values from a MODIS time series of a Normalized Difference Water Index (NDWI), using temporal weighting, was implemented to map persistent flooding in Louisiana coastal marshes. Ordinarily NDWI computed from daily 250m MODIS pixels represents a mixture of fragmented marshes and water. Here, sub-pixel NDWI values were derived for MODIS data using Landsat 30-m data. Each MODIS pixel was disaggregated into a mixture of the eight cover types according to the classified image pixels falling inside the MODIS pixel. The Landsat pixel means for each cover type inside a MODIS pixel were computed for the Landsat data preceding the MODIS image in time and for the Landsat data succeeding the MODIS image. The Landsat data were then weighted exponentially according to closeness in date to the MODIS data. The reconstructed MODIS data were produced by summing the product of fractional cover type with estimated NDWI values within each cover type. A new daily time series was produced using both the reconstructed 250-m MODIS, with enhanced features, and the approximated daily 30-m high-resolution image based on Landsat data. The algorithm was developed and tested over the Calcasieu-Sabine Basin, which was heavily inundated by storm surge from Hurricane Ike to study the extent and duration of flooding following the storm. Time series for 2000-2009, covering flooding events by Hurricane Rita in 2005 and Hurricane Ike in 2008, were derived. High resolution images were formed for all days in 2008 between the first cloud free Landsat scene and the last cloud-free Landsat scene. To refine and validate flooding maps, each time series was compared to Louisiana Coastwide Reference Monitoring System (CRMS) station water levels adjusted to marsh to optimize thresholds for MODIS-derived time series of NDWI. Seasonal fluctuations were adjusted by subtracting ten year average NDWI for marshes, excluding the hurricane events. Results from different NDWI indices and a combination of indices were compared. Flooding persistence that was mapped with higher-resolution data showed some improvement over the original MODIS time series estimates. The advantage of this novel technique is that improved mapping of extent and duration of inundation can be provided.

Underwood, L. W.↗
DOE OSTI2014

Smart trigger logic for focal plane arrays

An electronic device includes a memory configured to receive data representing light intensity values from pixels in a focal plane array and a processor that analyzes the received data to determine which light values correspond to triggered pixels, where the triggered pixels are those pixels that meet a predefined set of criteria, and determines, for each triggered pixel, a set of neighbor pixels for which light intensity values are to be stored. The electronic device also includes a buffer that temporarily stores light intensity values for at least one previously processed row of pixels, so that when a triggered pixel is identified in a current row, light intensity values for the neighbor pixels in the previously processed row and for the triggered pixel are persistently stored, as well as a data transmitter that transmits the persistently stored light intensity values for the triggered and neighbor pixels to a data receiver.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Texture analysis of small renal cell carcinomas at MDCT for predicting relevant histologic and protein biomarkers

Purpose: To assess CT texture features of small renal cell carcinomas (≤ 4cm) for association with key pathologic features including protein biomarkers. Methods: Quantitative CT texture analysis (CTTA) of small renal cancers (≤ 4cm) was performed on non-contrast and portal venous phase abdominal MDCT scans with an ROI drawn at the largest cross-sectional diameter of the tumor using commercially available software. Texture parameters including mean pixel attenuation, the standard deviation (SD) of the pixel distribution histogram, entropy, the mean of positive pixels, the skewness (i.e., asymmetry) of the pixel histogram, kurtosis (i.e., peakness) of the pixel histogram, and the percentage of positive pixels were correlated with pathologic data from surgical resection, including histology and nuclear grade, as well as microarray analysis in a subset (n = 40) including Ki67 index, CRP, and neovascularization (CD105/CD31). Results: Portal venous phase images were available in 249 patients (105 women, 144 men; mean age, 56.7 years) with tumors ≤ 4cm (mean, median, range, ± SD; 2.66, 2.60, 0.3–4.0 ± 0.85 cm). CT texture features of standard deviation, mean of the positive pixels, and entropy of the pixel histogram were significantly associated with histologic cell type (clear vs. non-clear; p < 0.001). Entropy and mean of the positive pixels also showed an association with nuclear grade, although not statistically significant. In the microarray analysis subset, kurtosis of the pixel histogram was associated with CD105/CD31 (p = 0.05). SD also showed some association with CD 105 positivity (p = 0.02) and CAIX expression (p = 0.01). Non-contrast CT images were available in 174 patients (72 women, 102 men; mean age, 57.5 years). Although the association with histology was not as strong as on the portal venous phase, in the subset of patients with microarray data, SD was found to correlate with CRP (p = 0.08), kurtosis with CRP (p = 0.004), CD105/CD31 (p = 0.002), and with Ki 67 index (p < 0.001). Conclusion: CT texture features were significantly associated with important histopathologic features in small renal cancers. These non-invasive measures can be performed retrospectively and may provide useful information when determining follow-up and treatment of small renal cancers.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2016

Indium-bump-free antimonide superlattice membrane detectors on silicon substrates

We present an approach to realize antimonide superlattices on silicon substrates without using conventional Indium-bump hybridization. In this approach, PIN superlattices are grown on top of a 60 nm Al0.6Ga0.4Sb sacrificial layer on a GaSb host substrate. Following the growth, the individual pixels are transferred using our epitaxial-lift off technique, which consists of a wet-etch to undercut the pixels followed by a dry-stamp process to transfer the pixels to a silicon substrate prepared with a gold layer. Structural and optical characterization of the transferred pixels was done using an optical microscope, scanning electron microscopy, and photoluminescence. The interface between the transferred pixels and the new substrate was abrupt, and no significant degradation in the optical quality was observed. An Indium-bump-free membrane detector was then fabricated using this approach. Spectral response measurements provided a 100% cut-off wavelength of 4.3 μm at 77 K. The performance of the membrane detector was compared to a control detector on the as-grown substrate. The membrane detector was limited by surface leakage current. The proposed approach could pave the way for wafer-level integration of photonic detectors on silicon substrates, which could dramatically reduce the cost of these detectors.

Zamiri, M. (ORCID:000000020042735X)↗
NASA NTRS1982

The digital step edge

The facet model was used to accomplish step edge detection. The essence of the facet model is that any analysis made on the basis of the pixel values in some neighborhood has its final authoritative interpretation relative to the underlying grey tone intensity surface of which the neighborhood pixel values are observed noisy samples. Pixels which are part of regions have simple grey tone intensity surfaces over their areas. Pixels which have an edge in them have complex grey tone intensity surfaces over their areas. Specially, an edge moves through a pixel only if there is some point in the pixel's area having a zero crossing of the second directional derivative taken in the direction of a non-zero gradient at the pixel's center. To determine whether or not a pixel should be marked as a step edge pixel, its underlying grey tone intensity surface was estimated on the basis of the pixels in its neighborhood.

Haralick, R. M.↗
NASA NTRS1990

Programmable remapper with single flow architecture

The invention relates to image processing systems and methods and in particular to a machine which accepts a real time video image in the form of a matrix of picture elements (pixels) and remaps such image according to a selectable one of a plurality of mapping functions to create an output matrix of pixels. Such mapping functions, or transformations, may be any one of a number of different transformations depending on the objective of the user of the system. The system remaps input images from one coordinate system to another using a set of look-up tables for the data necessary for the transform. The transforms, which are operator selectable, are precomputed and loaded into massive look-up tables. Input pixels, via the look-up tables of any particular transform selected, are mapped into output pixels with the radiance information of the input pixels being appropriately weighted. An earlier embodiment of the system included two parallel processors: a collective processor which mapped multiple input pixels into a single output pixel and an interpolative processor. The interpolative processor performed an interpolation among pixels in the input image where a given input pixel may affect the value of many output pixels. Several advantages are provided over previous embodiments in that the two distinct processors are replaced by a single processor capable of performing both types of operations (collective and interpolative) with no more complexity. Previously, there has existed no image processor or 'remapper' that can operate with sufficient speed and flexibility to permit investigating different transformation patterns in real time.

Fisher, Timothy E.↗
NASA NTRS2008

On Certain New Methodology for Reducing Sensor and Readout Electronics Circuitry Noise in Digital Domain

NASA Hubble Space Telescope (HST) and upcoming cosmology science missions carry instruments with multiple focal planes populated with many large sensor detector arrays. These sensors are passively cooled to low temperatures for low-level light (L3) and near-infrared (NIR) signal detection, and the sensor readout electronics circuitry must perform at extremely low noise levels to enable new required science measurements. Because we are at the technological edge of enhanced performance for sensors and readout electronics circuitry, as determined by thermal noise level at given temperature in analog domain, we must find new ways of further compensating for the noise in the signal digital domain. To facilitate this new approach, state-of-the-art sensors are augmented at their array hardware boundaries by non-illuminated reference pixels, which can be used to reduce noise attributed to sensors. There are a few proposed methodologies of processing in the digital domain the information carried by reference pixels, as employed by the Hubble Space Telescope and the James Webb Space Telescope Projects. These methods involve using spatial and temporal statistical parameters derived from boundary reference pixel information to enhance the active (non-reference) pixel signals. To make a step beyond this heritage methodology, we apply the NASA-developed technology known as the Hilbert- Huang Transform Data Processing System (HHT-DPS) for reference pixel information processing and its utilization in reconfigurable hardware on-board a spaceflight instrument or post-processing on the ground. The methodology examines signal processing for a 2-D domain, in which high-variance components of the thermal noise are carried by both active and reference pixels, similar to that in processing of low-voltage differential signals and subtraction of a single analog reference pixel from all active pixels on the sensor. Heritage methods using the aforementioned statistical parameters in the digital domain (such as statistical averaging of the reference pixels themselves) zeroes out the high-variance components, and the counterpart components in the active pixels remain uncorrected. This paper describes how the new methodology was demonstrated through analysis of fast-varying noise components using the Hilbert-Huang Transform Data Processing System tool (HHT-DPS) developed at NASA and the high-level programming language MATLAB (Trademark of MathWorks Inc.), as well as alternative methods for correcting for the high-variance noise component, using an HgCdTe sensor data. The NASA Hubble Space Telescope data post-processing, as well as future deep-space cosmology projects on-board instrument data processing from all the sensor channels, would benefit from this effort.

Kizhner, Semion↗
NASA NTRS2008

Chromatic Modulator for a High-Resolution CCD or APS

A chromatic modulator has been proposed to enable the separate detection of the red, green, and blue (RGB) color components of the same scene by a single charge-coupled device (CCD), active-pixel sensor (APS), or similar electronic image detector. Traditionally, the RGB color-separation problem in an electronic camera has been solved by use of either (1) fixed color filters over three separate image detectors; (2) a filter wheel that repeatedly imposes a red, then a green, then a blue filter over a single image detector; or (3) different fixed color filters over adjacent pixels. The use of separate image detectors necessitates precise registration of the detectors and the use of complicated optics; filter wheels are expensive and add considerably to the bulk of the camera; and fixed pixelated color filters reduce spatial resolution and introduce color-aliasing effects. The proposed chromatic modulator would not exhibit any of these shortcomings. The proposed chromatic modulator would be an electromechanical device fabricated by micromachining. It would include a filter having a spatially periodic pattern of RGB strips at a pitch equal to that of the pixels of the image detector. The filter would be placed in front of the image detector, supported at its periphery by a spring suspension and electrostatic comb drive. The spring suspension would bias the filter toward a middle position in which each filter strip would be registered with a row of pixels of the image detector. Hard stops would limit the excursion of the spring suspension to precisely one pixel row above and one pixel row below the middle position. In operation, the electrostatic comb drive would be actuated to repeatedly snap the filter to the upper extreme, middle, and lower extreme positions. This action would repeatedly place a succession of the differently colored filter strips in front of each pixel of the image detector. To simplify the processing, it would be desirable to encode information on the color of the filter strip over each row (or at least over some representative rows) of pixels at a given instant of time in synchronism with the pixel output at that instant.

Hartley, Frank↗
NASA NTRS2008

Model of Image Artifacts from Dust Particles

A mathematical model of image artifacts produced by dust particles on lenses has been derived. Machine-vision systems often have to work with camera lenses that become dusty during use. Dust particles on the front surface of a lens produce image artifacts that can potentially affect the performance of a machine-vision algorithm. The present model satisfies a need for a means of synthesizing dust image artifacts for testing machine-vision algorithms for robustness (or the lack thereof) in the presence of dust on lenses. A dust particle can absorb light or scatter light out of some pixels, thereby giving rise to a dark dust artifact. It can also scatter light into other pixels, thereby giving rise to a bright dust artifact. For the sake of simplicity, this model deals only with dark dust artifacts. The model effectively represents dark dust artifacts as an attenuation image consisting of an array of diffuse darkened spots centered at image locations corresponding to the locations of dust particles. The dust artifacts are computationally incorporated into a given test image by simply multiplying the brightness value of each pixel by a transmission factor that incorporates the factor of attenuation, by dust particles, of the light incident on that pixel. With respect to computation of the attenuation and transmission factors, the model is based on a first-order geometric (ray)-optics treatment of the shadows cast by dust particles on the image detector. In this model, the light collected by a pixel is deemed to be confined to a pair of cones defined by the location of the pixel s image in object space, the entrance pupil of the lens, and the location of the pixel in the image plane (see Figure 1). For simplicity, it is assumed that the size of a dust particle is somewhat less than the diameter, at the front surface of the lens, of any collection cone containing all or part of that dust particle. Under this assumption, the shape of any individual dust particle artifact is the shape (typically, circular) of the aperture, and the contribution of the particle to the attenuation factor for a given pixel is the fraction of the cross-sectional area of the collection cone occupied by the particle. Assuming that dust particles do not overlap, the net transmission factor for a given pixel is calculated as one minus the sum of attenuation factors contributed by all dust particles affecting that pixel. In a test, the model was used to synthesize attenuation images for random distributions of dust particles on the front surface of a lens at various relative aperture (F-number) settings. As shown in Figure 2, the attenuation images resembled dust artifacts in real test images recorded while the lens was aimed at a white target.

Willson, Reg↗
DOE OSTI2018

Reflective photovoltaics

A photovoltaic module includes colorized reflective photovoltaic cells that act as pixels. The colorized reflective photovoltaic cells are arranged so that reflections from the photovoltaic cells or pixels visually combine into an image on the photovoltaic module. The colorized photovoltaic cell or pixel is composed of a set of 100 to 256 base color sub-pixel reflective segments or sub-pixels. The color of each pixel is determined by the combination of base color sub-pixels forming the pixel. As a result, each pixel can have a wide variety of colors using a set of base colors, which are created, from sub-pixel reflective segments having standard film thicknesses.

14 SOLAR ENERGY↗
DOE OSTI2015

Fabrication of large dual-polarized multichroic TES bolometer arrays for CMB measurements with the SPT-3G camera

This paper presents the procedures used by Argonne National Laboratory to fabricate large arrays of multichroic transition-edge sensor (TES) bolometers for cosmic microwave background (CMB) measurements. These detectors will be assembled into the focal plane for the SPT-3G camera, the third generation CMB camera to be installed in the South Pole Telescope. The complete SPT-3G camera will have approximately 2690 pixels, for a total of 16,140 TES bolometric detectors. Each pixel is comprised of a broad-band sinuous antenna coupled to a Nb microstrip line. In-line filters are used to define the different band-passes before the millimeter-wavelength signal is fed to the respective Ti/Au TES bolometers. There are six TES bolometer detectors per pixel, which allow for measurements of three band-passes (95 GHz, 150 GHz and 220 GHz) and two polarizations. The steps involved in the monolithic fabrication of these detector arrays are presented here in detail. Patterns are defined using a combination of stepper and contact lithography. The misalignment between layers is kept below 200 nm. The overall fabrication involves a total of 16 processes, including reactive and magnetron sputtering, reactive ion etching, inductively coupled plasma etching and chemical etching.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2016

SU-C-206-03: Metal Artifact Reduction in X-Ray Computed Tomography Based On Local Anatomical Similarity

Purpose: Metal implants such as orthopedic hardware and dental fillings cause severe bright and dark streaking in reconstructed CT images. These artifacts decrease image contrast and degrade HU accuracy, leading to inaccuracies in target delineation and dose calculation. Additionally, such artifacts negatively impact patient set-up in image guided radiation therapy (IGRT). In this work, we propose a novel method for metal artifact reduction which utilizes the anatomical similarity between neighboring CT slices. Methods: Neighboring CT slices show similar anatomy. Based on this anatomical similarity, the proposed method replaces corrupted CT pixels with pixels from adjacent, artifact-free slices. A gamma map, which is the weighted summation of relative HU error and distance error, is calculated for each pixel in the artifact-corrupted CT image. The minimum value in each pixel’s gamma map is used to identify a pixel from the adjacent CT slice to replace the corresponding artifact-corrupted pixel. This replacement only occurs if the minimum value in a particular pixel’s gamma map is larger than a threshold. The proposed method was evaluated with clinical images. Results: Highly attenuating dental fillings and hip implants cause severe streaking artifacts on CT images. The proposed method eliminates the dark and bright streaking and improves the implant delineation and visibility. In particular, the image non-uniformity in the central region of interest was reduced from 1.88 and 1.01 to 0.28 and 0.35, respectively. Further, the mean CT HU error was reduced from 328 HU and 460 HU to 60 HU and 36 HU, respectively. Conclusions: The proposed metal artifact reduction method replaces corrupted image pixels with pixels from neighboring slices that are free of metal artifacts. This method proved capable of suppressing streaking artifacts, improving HU accuracy and image detectability.

60 APPLIED LIFE SCIENCES↗
NASA NTRS1995

Circuit for high resolution decoding of multi-anode microchannel array detectors

A circuit for high resolution decoding of multi-anode microchannel array detectors consisting of input registers accepting transient inputs from the anode array; anode encoding logic circuits connected to the input registers; midpoint pipeline registers connected to the anode encoding logic circuits; and pixel decoding logic circuits connected to the midpoint pipeline registers is described. A high resolution algorithm circuit operates in parallel with the pixel decoding logic circuit and computes a high resolution least significant bit to enhance the multianode microchannel array detector's spatial resolution by halving the pixel size and doubling the number of pixels in each axis of the anode array. A multiplexer is connected to the pixel decoding logic circuit and allows a user selectable pixel address output according to the actual multi-anode microchannel array detector anode array size. An output register concatenates the high resolution least significant bit onto the standard ten bit pixel address location to provide an eleven bit pixel address, and also stores the full eleven bit pixel address. A timing and control state machine is connected to the input registers, the anode encoding logic circuits, and the output register for managing the overall operation of the circuit.

Kasle, David B.↗
NASA NTRS1998

Does the UVI on Polar Detect Cosmic Snowballs?

Frank and Sigwarth [1997a] claim that the dark pixels observed in dayglow images obtained by the Earth sensor of the Visible Imaging System (VIS) are due to bombardment of Earth by 20 to 40 ton cosmic snowballs. We have independently studied the same one hour of VIS data Frank and Sigwarth used and have performed detailed statistical analysis of the dark pixels. The characteristics of the dark pixels from the VIS images have been compared to those obtained from the overlapping images from the Ultraviolet Imager (UVI). We find the occurrence distributions of the dark pixels, single and multiple, from VIS and UVI are nearly identical. This result cannot be explained by a -eophysical source since the two cameras have different pixel resolutions: A search for evidence of of spacecraft "wobble" motion, whose presence would indicate that the source is external to the camera, has found that pairs of dark pixel clusters are uniformly distributed in an-le and no preference is observed in the wobble direction. Instrument artifacts as the source of the dark pixels is the most likely explanation for these results. Probability estimates for the occurrence of dark pixel clusters lead us to expect coincident events of instrumental origin to occur frequently in the two cameras. The conclusion of this study is that neither VIS nor UVI provide any scientific evidence that the origin of dark pixels is geophysical.

Parks, G. K.↗
NASA NTRS2007

[Overview of the GSFC Effort to Build Microwave Detectors for CMB Polarization]

Filled arrays of bolometers are currently being employed for use in astronomy from the far-infrared through millimeter portion of the electromagnetic spectrum. Because of the large range of wavelengths for which such detectors are applicable, the size of a pixel (p) relative to the wavelength (lambda) will vary according to the specific application of a given available technology. As the pixel size becomes comparable to a wavelength of the absorbed radiation, correlations in the radiation field at the pixel can affect the imaging properties of the system independent of the correlations introduced by the front-end optics. We study the dependence of image fidelity and induced polarization on the size of the pixel by employing a formalism in which diffraction due to the pixel boundary is treated by propagating the second-order statistical correlations of the radiation field through a model optical system. We construct simulated polarized images of square pixels for various ratios of p lambda. For the limit in which few modes are supported by the pixel (p lambda lesssiml), we find that the diffraction due to the pixel edges is non-negligible and hence must be considered along with the telescope diffraction pattern in modeling the ultimate spatial resolution of an imaging system. For the case in which the pixel is over-moded (p lambda ggl), the geometric limit is approached as expected. This technique provides a quantitative approach to understand and optimize the imaging properties of planar detectors in the few-mode limit.

Chuss, David↗
NASA NTRS2010

A Framework for Propagation of Uncertainties in the Kepler Data Analysis Pipeline

The Kepler space telescope is designed to detect Earth-like planets around Sun-like stars using transit photometry by simultaneously observing 100,000 stellar targets nearly continuously over a three and a half year period. The 96-megapixel focal plane consists of 42 charge-coupled devices (CCD) each containing two 1024 x 1100 pixel arrays. Cross-correlations between calibrated pixels are introduced by common calibrations performed on each CCD requiring downstream data products access to the calibrated pixel covariance matrix in order to properly estimate uncertainties. The prohibitively large covariance matrices corresponding to the ~75,000 calibrated pixels per CCD preclude calculating and storing the covariance in standard lock-step fashion. We present a novel framework used to implement standard propagation of uncertainties (POU) in the Kepler Science Operations Center (SOC) data processing pipeline. The POU framework captures the variance of the raw pixel data and the kernel of each subsequent calibration transformation allowing the full covariance matrix of any subset of calibrated pixels to be recalled on-the-fly at any step in the calibration process. Singular value decomposition (SVD) is used to compress and low-pass filter the raw uncertainty data as well as any data dependent kernels. The combination of POU framework and SVD compression provide downstream consumers of the calibrated pixel data access to the full covariance matrix of any subset of the calibrated pixels traceable to pixel level measurement uncertainties without having to store, retrieve and operate on prohibitively large covariance matrices. We describe the POU Framework and SVD compression scheme and its implementation in the Kepler SOC pipeline.

Clarke, Bruce D.↗
NASA NTRS2011

Method and system for non-linear motion estimation

A method and system for extrapolating and interpolating a visual signal including determining a first motion vector between a first pixel position in a first image to a second pixel position in a second image, determining a second motion vector between the second pixel position in the second image and a third pixel position in a third image, determining a third motion vector between one of the first pixel position in the first image and the second pixel position in the second image, and the second pixel position in the second image and the third pixel position in the third image using a non-linear model, determining a position of the fourth pixel in a fourth image based upon the third motion vector.

Lu, Ligang↗
NASA NTRS2011

Superpixel-Augmented Endmember Detection for Hyperspectral Images

Superpixels are homogeneous image regions comprised of several contiguous pixels. They are produced by shattering the image into contiguous, homogeneous regions that each cover between 20 and 100 image pixels. The segmentation aims for a many-to-one mapping from superpixels to image features; each image feature could contain several superpixels, but each superpixel occupies no more than one image feature. This conservative segmentation is relatively easy to automate in a robust fashion. Superpixel processing is related to the more general idea of improving hyperspectral analysis through spatial constraints, which can recognize subtle features at or below the level of noise by exploiting the fact that their spectral signatures are found in neighboring pixels. Recent work has explored spatial constraints for endmember extraction, showing significant advantages over techniques that ignore pixels relative positions. Methods such as AMEE (automated morphological endmember extraction) express spatial influence using fixed isometric relationships a local square window or Euclidean distance in pixel coordinates. In other words, two pixels covariances are based on their spatial proximity, but are independent of their absolute location in the scene. These isometric spatial constraints are most appropriate when spectral variation is smooth and constant over the image. Superpixels are simple to implement, efficient to compute, and are empirically effective. They can be used as a preprocessing step with any desired endmember extraction technique. Superpixels also have a solid theoretical basis in the hyperspectral linear mixing model, making them a principled approach for improving endmember extraction. Unlike existing approaches, superpixels can accommodate non-isometric covariance between image pixels (characteristic of discrete image features separated by step discontinuities). These kinds of image features are common in natural scenes. Analysts can substitute superpixels for image pixels during endmember analysis that leverages the spatial contiguity of scene features to enhance subtle spectral features. Superpixels define populations of image pixels that are independent samples from each image feature, permitting robust estimation of spectral properties, and reducing measurement noise in proportion to the area of the superpixel. This permits improved endmember extraction, and enables automated search for novel and constituent minerals in very noisy, hyperspatial images. This innovation begins with a graph-based segmentation based on the work of Felzenszwalb et al., but then expands their approach to the hyperspectral image domain with a Euclidean distance metric. Then, the mean spectrum of each segment is computed, and the resulting data cloud is used as input into sequential maximum angle convex cone (SMACC) endmember extraction.

Thompson, David R.↗