Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “centroid estimation”

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 73 records · Page 4
All sourcesNASA NTRSDOE OSTI
NASA NTRS2007

Optimal Calibration of the Spitzer Space Telescope

A document discusses the focal-plane calibration of the Spitzer Space Telescope by use of the instrument pointing frame (IPF) Kalman filter, which was described in Kalman Filter for Calibrating a Telescope Focal Plane (NPO-40798), NASA Tech Briefs, Vol. 30, No. 9 (September 2006), page 62. To recapitulate: In the IPF Kalman filter, optimal estimates of both engineering and scientific focal-plane parameters are obtained simultaneously, using data taken in each focalplane survey activity. The IPF Kalman filter offers greater efficiency and economy, relative to prior calibration practice in which scientific and engineering parameters were estimated by separate teams of scientists and engineers and iterated upon each other. In the Spitzer Space Telescope application, the IPF Kalman filter was used to calibrate 56 frames for precise telescope pointing, estimate >1,500 parameters associated with focal-plane mapping, and process calibration runs involving as many as 1,338 scientific image centroids. The final typical survey calibration accuracy was found to be 0.09 arc second. The use of the IPF Kalman filter enabled a team of only four analysts to complete the calibration processing in three months. An unanticipated benefit afforded by the IPF Kalman filter was the ability to monitor health and diagnose performance of the entire end-to-end telescope-pointing system.

Bayard, David↗
NASA NTRS2019

Camera modeling, centroiding performance, and geometric camera calibration on ASTERIA

The Arcsecond Space Telescope Enabling Research in Astrophysics (ASTERIA) is a 10-kg, 6U CubeSat in low-Earth orbit that was able to achieve subarcsecond pointing stability and repeatability. To date, this is the best pointing on a spacecraft of its size. This paper will analyze various aspects of the performance of its key piece of hardware—the payload. First,a model of the optics and imager, which is used to simulate stellar images, will be presented. The imager parameters used in this model were derived from simple ground measurements. Next, a centroiding algorithm is provided and used on the simulated images to predict centroiding performance. These results will be shown to match on-orbit telemetry of centroiding performance, validating the modeling approach. This paper will then describe an approach for and results of a geometric camera calibration algorithm to estimate the focal length,distortion,and alignment parameters. The modeling, analyses, and results presented in this paper provide key information that can be used in a time domain pointing simulation or a frequency-domain pointing error analysis.

Smith, Matthew W.↗
NASA NTRS2012

The Height of a White-Light Flare and its Hard X-Ray Sources

We describe observations of a white-light (WL) flare (SOL2011-02-24T07:35:00, M3.5) close to the limb of the Sun, from which we obtain estimates of the heights of the optical continuum sources and those of the associated hard X-ray (HXR) sources. For this purpose, we use HXR images from the Reuven Ramaty High Energy Spectroscopic Imager and optical images at 6173 Ang. from the Solar Dynamics Observatory.We find that the centroids of the impulsive-phase emissions in WL and HXRs (30 -80 keV) match closely in central distance (angular displacement from Sun center), within uncertainties of order 0".2. This directly implies a common source height for these radiations, strengthening the connection between visible flare continuum formation and the accelerated electrons. We also estimate the absolute heights of these emissions as vertical distances from Sun center. Such a direct estimation has not been done previously, to our knowledge. Using a simultaneous 195 Ang. image from the Solar-Terrestrial RElations Observatory spacecraft to identify the heliographic coordinates of the flare footpoints, we determine mean heights above the photosphere (as normally defined; tau = 1 at 5000 Ang.) of 305 +/- 170 km and 195 +/- 70 km, respectively, for the centroids of the HXR and WL footpoint sources of the flare. These heights are unexpectedly low in the atmosphere, and are consistent with the expected locations of tau = 1 for the 6173 Ang and the approx 40 keV photons observed, respectively.

Sun: photosphere↗
NASA NTRS2012

Top-of-the-Atmosphere Shortwave Flux Estimation from UV Observations: An Empirical Approach

Measurements of top of the atmosphere (TOA) radiation are essential to the understanding of Earth's climate. Clouds, aerosols, and ozone (0,) are among the most important agents impacting the Earth's short-wave (SW) radiation budget. There are several sensors in orbit that provide independent information related to the Earth's SW radiation budget. Having coincident information from these sensors is important for understanding their potential contributions. The A-train constellation of satellites provides a unique opportunity to analyze near-simultaneous data from several of these sensors. They include the Ozone Monitoring Instrument (OMI), on the NASA Aura satellite, that makes TOA hyper-spectral measurements from ultraviolet (UV) to visible wavelengths, and Clouds and the Earth's Radiant Energy System (CERES) instrument, on the NASA Aqua satellite, that makes broadband measurements in both the long- and short-wave. OMI measurements have been successfully utilized to derive the information on trace gases (e.g., 0 1, NO" and SO,), clouds, and absorbing aerosols. TOA SW fluxes are estimated using a combination of data from CERES and the Aqua MODerate-resolution Imaging Spectroradiometer (MODIS). In this paper, OMI retrievals of cloud/aerosol parameters and 0 1 have been collocated with CERES TOA SW flux retrievals. We use this collocated data to develop a neural network that estimates TOA shortwave flux globally over ocean using data from OMI and meteorological analyses. This input data include the effective cloud fraction, cloud optical centroid pressure (OCP), total-column 0" and sun-satellite viewing geometry from OMI as well as wind speed and water vapor from the Goddard Earth Observing System 5 Modern Era Retrospective-analysis for Research and Applications (GEOS-5 MERRA) along with a climatology of chlorophyll content. We train the neural network using a subset of CERES retrievals of TOA SW flux as the target output (truth) and withhold a different subset of the CERES data to be used for validation.

Gupta, P.↗
NASA NTRS2005

Thermal Imaging of Earth for Accurate Pointing of Deep-Space Antennas

A report discusses a proposal to use thermal (long-wavelength infrared) images of the Earth, as seen from spacecraft at interplanetary distances, for pointing antennas and telescopes toward the Earth for Ka-band and optical communications. The purpose is to overcome two limitations of using visible images: (1) at large Earth phase angles, the light from the Earth is too faint; and (2) performance is degraded by large albedo variations associated with weather changes. In particular, it is proposed to use images in the wavelength band of 8 to 13 m, wherein the appearance of the Earth is substantially independent of the Earth phase angle and emissivity variations are small. The report addresses tracking requirements for optical and Ka-band communications, selection of the wavelength band, available signal level versus phase angle, background noise, and signal-to-noise ratio. Tracking errors are estimated for several conceptual systems employing currently available infrared image sensors. It is found that at Mars range, it should be possible to locate the centroid of the Earth image within a noise equivalent angle (a random angular error) between 10 and 150 nanoradians at a bias error of no more than 80 nanoradians

Ortiz, Gerardo↗
NASA NTRS2016

Unsteady Aerodynamic Force Sensing from Measured Strain

A simple approach for computing unsteady aerodynamic forces from simulated measured strain data is proposed in this study. First, the deflection and slope of the structure are computed from the unsteady strain using the two-step approach. Velocities and accelerations of the structure are computed using the autoregressive moving average model, on-line parameter estimator, low-pass filter, and a least-squares curve fitting method together with analytical derivatives with respect to time. Finally, aerodynamic forces over the wing are computed using modal aerodynamic influence coefficient matrices, a rational function approximation, and a time-marching algorithm. A cantilevered rectangular wing built and tested at the NASA Langley Research Center (Hampton, Virginia, USA) in 1959 is used to validate the simple approach. Unsteady aerodynamic forces as well as wing deflections, velocities, accelerations, and strains are computed using the CFL3D computational fluid dynamics (CFD) code and an MSC/NASTRAN code (MSC Software Corporation, Newport Beach, California, USA), and these CFL3D-based results are assumed as measured quantities. Based on the measured strains, wing deflections, velocities, accelerations, and aerodynamic forces are computed using the proposed approach. These computed deflections, velocities, accelerations, and unsteady aerodynamic forces are compared with the CFL3D/NASTRAN-based results. In general, computed aerodynamic forces based on the lifting surface theory in subsonic speeds are in good agreement with the target aerodynamic forces generated using CFL3D code with the Euler equation. Excellent aeroelastic responses are obtained even with unsteady strain data under the signal to noise ratio of -9.8dB. The deflections, velocities, and accelerations at each sensor location are independent of structural and aerodynamic models. Therefore, the distributed strain data together with the current proposed approaches can be used as distributed deflection, velocity, and acceleration sensors. This research demonstrates the feasibility of obtaining induced drag and lift forces through the use of distributed sensor technology with measured strain data. An active induced drag control system thus can be designed using the two computed aerodynamic forces, induced drag and lift, to improve the fuel efficiency of an aircraft. Interpolation elements between structural finite element grids and the CFD grids and centroids are successfully incorporated with the unsteady aeroelastic computation scheme. The most critical technology for the success of the proposed approach is the robust on-line parameter estimator, since the least-squares curve fitting method depends heavily on aeroelastic system frequencies and damping factors.

shape sensing↗
NASA NTRS1987

Earthquake source mechanisms and transform fault tectonics in the Gulf of California

The source parameters of 19 earthquakes in the Gulf of California were determined from an inversion of long-period P and SH waveforms. The inversion procedure is described and the estimated precision of the derived source parameters is examined, with particular attention given to source complexity, the resolution of slip vector azimuth, and the resolution of centroid depth. The implications of these earthquake source characteristics for the tectonic evolution of the gulf are discussed.

Goff, John A.↗
DOE OSTI2026

Bayesian Exploration and Surrogate Emulation of Nonlinear Beam-Response Geometry in the LBNF Beamline

Next-generation long-baseline neutrino experiments aim to achieve multi-MW proton beam power while reducing accelerator-induced systematic uncertainties. At Fermilab, the LBNF beamline is designed for 1.2 MW operation with PIP-II and is upgradeable to 2.4 MW. DUNE will probe the three-flavor neutrino paradigm and search for CP violation, requiring precise neutrino-flux normalization and improved control of accelerator-related uncertainties. Within the LBNF beamline, the System for On-Axis Neutrino Detection (SAND) will constrain flux uncertainties using precision near-detector measurements, while the Muon Monitor System (MuMS) will provide beamline diagnostics sensitive to the proton beam, target, and horn configuration. However, the pion phase space relevant for DUNE depends simultaneously on many correlated parameters, including beam centroid, beam width, horn current and alignment, target position, optics shifts, and radiation-induced changes. Consequently, MuMS observables exhibit nonlinear and coupled responses that are difficult to characterize using traditional one-parameter scans. To address this challenge, we are developing a Bayesian Exploration framework coupled to physics-informed surrogate emulators trained on Geant4 beamline simulations. Gaussian-process emulators provide both fast predictions and uncertainty estimates, enabling adaptive selection of new simulation points in beam-parameter space. As an initial demonstration, we construct surrogate emulators for MuMS response observables using a verified simulation campaign spanning proton-beam steering conditions. The emulators reproduce the simulated dependence of MuMS centroid and gradient observables while providing predictive uncertainties, and serve as the foundation for future multidimensional exploration including beam width, horn current, and additional beamline parameters. This work establishes a framework for uncertainty-aware beam monitoring, adaptive simulation campaigns, and rapid beam-response inference for future DUNE operations.

Ganguly, Sudeshna [Fermilab] (ORCID:00000003163482↗
NASA NTRS1988

An orbit simulation study of a geopotential research mission including satellite-to-satellite tracking and disturbance compensation systems

Two orbit simulations, one representing the actual Geopotential Research Mission (GRM) orbit and the other representing the orbit estimated from orbit determination techniques, are presented. A computer algorithm was created to simulate GRM's drag compensation mechanism so the fuel expenditure and proof mass trajectories relative to the spacecraft centroid could be calculated for the mission. The results of the GRM DISCOS simulation demonstrated that the spacecraft can essentially be drag-free. The results showed that the centroid of the spacecraft can be controlled so that it will not deviate more than 1.0 mm in any direction from the centroid of the proof mass.

Antreasian, Peter G.↗
NASA NTRS2004

Determining Sizes of Particles in a Flow from DPIV Data

A proposed method of measuring the size of particles entrained in a flow of a liquid or gas would involve utilization of data from digital particle-image velocimetry (DPIV) of the flow. That is to say, with proper design and operation of a DPIV system, the DPIV data could be processed according to the proposed method to obtain particle sizes in addition to particle velocities. As an additional benefit, one could then compute the mass flux of the entrained particles from the particle sizes and velocities. As in DPIV as practiced heretofore, a pulsed laser beam would be formed into a thin sheet to illuminate a plane of interest in a flow field and the illuminated plane would be observed by means of a charge-coupled device (CCD) camera aimed along a line perpendicular to the illuminated plane. Unlike in DPIV as practiced heretofore, care would be taken to polarize the laser beam so that its electric field would lie in the illuminated plane, for the reason explained in the next paragraph. The proposed method applies, more specifically, to transparent or semitransparent spherical particles that have an index of refraction different from that of the fluid in which they are entrained. The method is based on the established Mie theory, which describes the scattering of light by diffraction, refraction, and specular reflection of light by such particles. In the case of a particle illuminated by polarized light and observed in the arrangement described in the preceding paragraph, the Mie theory shows that the image of the particle on the focal plane of the CCD camera includes two glare spots: one attributable to light reflected toward the camera and one attributable to light refracted toward the camera. The distance between the glare spots is a known function of the size of the particle, the indices of refraction of the particle material, and design parameters of the camera optics. Hence, the size of a particle can be determined from the distance between the glare spots. The proposed method would be implemented in an algorithm that would automatically identify, and measure the distance between, the glare spots for each particle for which a suitable image has been captured in a DPIV image frame. The algorithm (see figure) would begin with thresholding of data from the entire image frame to reduce noise, thereby facilitating discrimination of particle images from the background and aiding in the separation of overlapping particles. It is important not to pick a threshold level so high that the light intensity between a given pair of glare spots does not fall below the threshold value, leaving the glare spots disconnected. The image would then be scanned in a sequence of rows and columns of pixels to identify groups of adjacent pixels that contain nonzero brightnesses and that are surrounded by pixels of zero brightness. Each such group would be assumed to constitute the image of one particle. Each such group would be further analyzed to determine whether the image was saturated; saturated particle images must be rejected because the locations of glare spots in saturated images cannot accurately be determined. Within each unsaturated particle image, the centroids (deemed to be the locations) of the glare spots would be determined by means of gradients of brightness distributions and three-point horizontal and three-point vertical Gaussian estimates based on the brightness values of the brightest pixels and the pixels adjacent to them. If the brightness of a given particle image contained only one peak, then it would be assumed that a second glare spot did not exist and that image would be rejected.

Wernet, M. P.↗
NASA NTRS2013

On the Nature of the mHz X-ray Quasi-Periodic Oscillations from Ultraluminous X-ray source M82 X−1: Search for Timing-Spectral Correlations

Using all the archival XMM-Newton X-ray (3-10 keV) observations of the ultraluminous X-ray source (ULX) M82 X−1, we searched for a correlation between its variable mHz quasi-periodic oscillation (QPO) frequency and its hardness ratio (5-10 keV/3-5 keV), an indicator of the energy spectral power-law index. When stellar-mass black holes (StMBHs) exhibit type-C low-frequency QPOs (∼0.2-15 Hz), the centroid frequency of the QPO is known to correlate with the energy spectral index. The detection of such a correlation would strengthen the identification of M82 X−1's mHz QPOs as type-C and enable a more reliable mass estimate by scaling its QPO frequencies to those of type-C QPOs in StMBHs of known mass.We resolved the count rates and the hardness ratios of M82 X−1 and a nearby bright ULX (source 5/X42.3+59) through surface brightness modeling.We detected QPOs in the frequency range of 36-210 mHz during which M82 X−1's hardness ratio varied from 0.42 to 0.47. Our primary results are (1) that we do not detect any correlation between the mHz QPO frequency and the hardness ratio (a substitute for the energy spectral power-law index) and (2) similar to some accreting X-ray binaries, we find that M82 X−1's mHz QPO frequency increases with its X-ray count rate (Pearson's correlation coefficient = +0.97). The apparent lack of a correlation between the QPO centroid frequency and the hardness ratio poses a challenge to the earlier claims that the mHz QPOs of M82 X−1 are the analogs of the type-C low-frequency QPOs of StMBHs. On the other hand, it is possible that the observed relation between the hardness ratio and the QPO frequency represents the saturated portion of the correlation seen in type-C QPOs of StMBHs-in which case M82 X−1's mHz QPOs can still be analogous to type-C QPOs.

Nature↗
NASA NTRS2022

Properties of EUV Imaging Spectrometer (EIS) Slot Observations

The Extreme ultraviolet Imaging Spectrometer (EIS) on board the Hinode spacecraft has been operating since 2006, returning high-resolution data in the 170 – 212 and 246 – 292 Å wavelength regions. EIS has four slit options, with the narrow 1′′ and 2′′ slits used for spectroscopy and the wide 40′′ and 266′′ slits used for monochromatic imaging. In this article several properties of the 40′′ slit (or slot) are measured using the Fe XII 195.12 Å line, which is formed at 1.5 MK. The projected width of the slot on the detector shows a small variation along the slit with an average value of 40.949′′. The slot image is tilted on the detector and a quadratic formula is provided to describe the tilt. The tilt corresponds to four pixels on the detector and the slot centroid is offset mostly to the right (longer wavelengths) of the 1′′ slit by up to four pixels. Measurement of the intensity decrease at the edge of the slot leads to an estimate of the spatial resolution of the images in the x-direction. The resolution varies quadratically along the slot, with a minimum value of 2.9′′ close to the detector center. Intensities measured from the slot images are found to be on average 14% higher than those measured from the 1′′ slit at the same spatial location. Background subtraction is necessary to derive accurate intensities in quiet-Sun and coronal-hole regions. Prescriptions for deriving accurate slot intensities for different types of slot datasets are presented.

corona↗
NASA NTRS2018

Orion Optical Navigation Progress Toward Exploration: Mission 1

Optical navigation of human spacecraft was proposed on Gemini and implemented successfully on Apollo as a means of autonomously operating the vehicle in the event of lost communication with controllers on Earth. It shares a history with the "method of lunar distances" that was used in the 18th century and gained some notoriety after its use by Captain James Cook during his 1768 Pacific voyage of the HMS Endeavor. The Orion emergency return system utilizing optical navigation has matured in design over the last several years, and is currently undergoing the final implementation and test phase in preparation for Exploration Mission 1 (EM-1) in 2019. The software development is being worked as a Government Furnished Equipment (GFE) project delivered as an application within the Core Flight Software of the Orion camera controller module. The mathematical formulation behind the initial ellipse fit in the image processing is detailed in Christian. The non-linear least squares refinement then follows the technique of Mortari as an estimation process of the planetary limb using the sigmoid function. The Orion optical navigation system uses a body fixed camera, a decision that was driven by mass and mechanism constraints. The general concept of operations involves a 2-hour pass once every 24 hours, with passes specifically placed before all maneuvers to supply accurate navigation information to guidance and targeting. The pass lengths are limited by thermal constraints on the vehicle since the OpNav attitude generally deviates from the thermally stable tail-to-sun attitude maintained during the rest of the orbit coast phase. Calibration is scheduled prior to every pass due to the unknown nature of thermal effects on the lens distortion and the mounting platform deformations between the camera and star trackers. The calibration technique is described in detail by Christian, et al. and simultaneously estimates the Brown-Conrady coefficients and the Star Tracker/Camera interlock angles. Accurate attitude information is provided by the star trackers during each pass. Figure 1 shows the various phases of lunar return navigation when the vehicle is in autonomous operation with lost ground communication. The midcourse maneuvers are placed to control the entry interface conditions to the desired corridor for safe landing. The general form of optical navigation on Orion is where still images of the Moon or Earth are processed to find the apparent angular diameter and centroid in the camera focal plane. This raw data is transformed into range and bearing angle measurements using planetary data and precise star tracker inertial attitude. The measurements are then sent to the main flight computer's Kalman filter to update the onboard state vector. The images are, of course, collected over an arc to converge the state and estimate velocity. The same basic technique was used by Apollo to satisfy loss-of-comm, but Apollo used manual crew sightings with a vehicle-integral sextant instead of autonomously processing optical imagery. The software development is past its Critical Design Review, and is progressing through test and certification for human rating. In support of this, a hardware-in-the-loop test rig was developed in the Johnson Space Center Electro-Optics Lab to exercise the OpNav system prior to integrated testing on the Orion vehicle. Figure 2 shows the rig, which the test team has dubbed OCILOT (Orion Camera In the Loop Optical Testbed). Analysis performed to date shows a delivery that satisfies an allowable entry corridor as shown in Figure 3.

Holt, Greg N.↗
NASA NTRS1988

A stereo triangulation system for structural identification: Analytical and experimental results

Identification of large space structures' distributed mass, stiffness, and energy dissipation characteristics poses formidable analytical, numerical, and implementation difficulties. Development of reliable on-orbit structural identification methods is important for implementing active vibration suppression concepts which are under widespread study in the large space structures community. Near the heart of the identification problem lies the necessity of making a large number of spatially distributed measurements of the structure's vibratory response and the associated force/moment inputs with sufficient spatial and frequency resolution. In the present paper, we discuss a method whereby tens of active or passive (retro-reflecting) targets on the structure are tracked simultaneously by the focal planes of two or more video cameras mounted on an adjacent platform. Triangulation (optical ray intersection) of the conjugate image centroids yield inertial trajectories of each target on the structure. Given the triangulated motion of the targets, we apply and extend methodology developed by Creamer, Junkins, and Juang to identify the frequencies, mode shapes, and updated estimates for the mass/stiffness/damping parameterization of the structure. The methodology is semi-automated, for example, the post experiment analysis of the video imagery to determine the inertial trajectories of the targets typically requires less than thirty minutes of real time. Using methodology discussed herein, the frequency response of a large number of points on the structure (where reflective targets are mounted) on the structure can be determined from optical measurements alone. For comparison purposes, we also utilize measurements from accelerometers and a calibrated impulse hammer. While our experimental work remains in a research stage of development, we have successfully tracked and stereo triangulated 20 targets (on a vibrating cantilevered grid structure) at a sample frequency of 200 HZ, and have established conclusively the feasibility and desirability of this approach. We discuss, in summary, recent advances in analog and digital video processing methodology, actuation methods, and bring them to bear on the structural identification problem. We include a brief discussion of our experimental hardware and some recent experimental results which support the practical feasibility of this structural vibration sensing approach.

Junkins, J. L.↗
DOE OSTI2025

5-minute Wind Power Data based on WFIP2 WRF Simulation

The second Wind Forecast Improvement Project (WFIP2) was a public-private partnership funded by the U.S. Department of Energy and NOAA, aimed at enhancing the forecast skill of numerical weather prediction models for turbine-height winds in regions with complex terrain. An 18-month Weather Research and Forecasting (WRF) model simulation was conducted over the Pacific Northwest, with model outputs validated against observational data collected during WFIP2. Simulated wind speeds were used to estimate wind power generation using reV (the Renewable Energy Potential Model developed by NREL) at ten wind project sites. Two sets of results were produced: one using wind speeds extracted from the model grid cell at the project centroid, and another using wind speeds from the actual turbine locations. For each dataset, power output was calculated using both actual turbine-specific power curves and nine generic power curves to convert wind speed into power.

17 WIND ENERGY↗
NASA NTRS1999

A Diagnostic Analysis of the Kennedy Space Center LDAR Network: Data Characteristics - 1

An analytical framework is developed in which to analyze climatological VHF (66 MHz) radiation measurements taken by the Kennedy Space Center Lightning Detection and Ranging (LDAR) network. A 19-month noise-filtered sample of LDAR observations is examined using this framework. It is found that the climatological VHF source density as observed by LDAR falls off approximately 10 dB every 71 km of ground range away from the network centroid (a 31 km e-folding scale). From this framework it is inferred that the underlying source distribution is likely inverse exponential in amplitude, with a log-slope of 126-147 mW(sup -1/2). The underlying vertical distribution of VHF sources is approximately normally distributed with a mean altitude of 9 km and a standard deviation of 2.7 km; this implies that the loss of below-horizon sources has a negligible effect on column-integrated source densities within 200 km ground range. At medium to far ranges, location errors are primarily radial and have a slightly asymmetric distribution whose first moment increases as range squared. Error moments estimated from observed lightning are significantly higher than those from aircraft-based signal generator or analytic solution estimates, suggesting that timing errors arising from poor signal identification and discrimination may dominate over timing errors arising from nominal sensor resolution. The VHF source properties of individual LDAR-observed flashes are computed and an analytic expression for flash detection efficiency vs. range is derived. This reveals nearly constant flash detection efficiency to 80-94 km range from the network centroid.

Boccippio, D. J.↗
DOE OSTI2026

Toward an AI-Powered Software Pipeline for Real-Time Tracking and Analysis of Wildfire and Smoke

Real-time tracking of wildfires and smoke is crucial for effective response, minimizing damage, protecting lives, and efficiently managing resources during fire emergencies. We develop a web-based AI-powered pipeline that detects wildfires in aerial video and estimates deployment-relevant behavior metrics, including cumulative burned area, burned-area growth rate, fire spread direction, and smoke dispersion. The system combines a YOLO-based detector with YCbCr-based fire segmentation, HSV-based smoke segmentation, Farneback optical flow, and centroid-based spatiotemporal tracking. Using ground sampling distance (GSD), pixel-level fire masks are converted to physical burned-area measurements by correlating fire pixel counts with camera altitude and tilt angle. We benchmark YOLO variants and non-YOLO baselines (GoogLeNet, CNN, DBN, Autoencoder, U-Net, and AlexNet) on the IEEE FLAME dataset and a newly created aerial frame dataset, Wildfire-DB. Cross-dataset evaluation uses a strict threshold-transfer protocol: decision thresholds are selected on FLAME validation and transferred unchanged to Wildfire-DB to quantify generalization under domain shift. YOLOv6 achieves the strongest cross-dataset frame-level fire detection on Wildfire-DB (ROC-AUC 0.8200, PR-AUC 0.8044, and transferred-threshold F1 0.7596). For tracking-oriented deployment requiring oriented localization, YOLO11-OBB provides the most reliable cross-dataset behavior among OBB-capable models while remaining computationally feasible. To analyze the feasibility of UAV deployment, we further measure inference efficiency using synchronized GPU and CPU power logs on a fixed workload of 1569 frames. YOLO-family models process the video in 5.73–12.47 seconds with net energy of 1247.28–1775.39 J, substantially lower latency and energy than heavier classification and reconstruction baselines. Overall, model optimality depends on operational objectives: YOLOv6 is best for cross-dataset detection robustness, whereas YOL...

Color segmentation↗
DOE OSTI2026

Novel soft x-ray multi-energy camera to study thermal plasmas on WEST

A multi-energy soft x-ray (SXR) pinhole camera has been built, calibrated and deployed on the WEST tokamak to aid the study of thermal plasmas. This novel imaging diagnostic technique employs a 2D pixelated x-ray detector in which the lower energy threshold for photon detection can be adjusted independently on each pixel. This innovation allows the simultaneous measurement of x-rays with space, time and energy resolution. Measurements at lower energy (2–6 keV) dominated by tungsten line emission are used to estimate the tungsten concentration and the plasma ΔZ eff . Measurements at higher energy (11–18 keV), where only continuum emission is present, are employed to infer the plasma electron temperature. Other applications of this powerful diagnostic include the determination of the plasma SXR centroid vertical position, characterization of tungsten UFOs and various real-time measurements.

electron temperature↗