Search NASA⌕ Search

SEARCH · Search NASA

Results for “superresolution”

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 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Physics-constrained superresolution diffusion for six-dimensional phase space diagnostics

Adaptive physics-constrained superresolution diffusion is developed for noninvasive virtual diagnostics of the six-dimensional (6D) phase space density of charged particle beams. An adaptive variational autoencoder embeds initial beam condition images and scalar measurements to a low-dimensional latent space from which a 32 6 pixel 6D tensor representation of the beam's 6D phase space density is generated. Projecting from a 6D tensor generates physically consistent two-dimensional projections. Physics-guided superresolution diffusion transforms low-resolution images of the 6D density to high resolution 256 × 256 pixel images. Unsupervised adaptive latent space tuning enables tracking of time-varying beams without knowledge of time-varying initial conditions. The method is demonstrated with experimental data and multiparticle simulations at the HiRES UED. The general approach is applicable to a wide range of complex dynamic systems evolving in high-dimensional phase space. The method is shown to be robust to distribution shift without retraining. Published by the American Physical Society 2025

43 PARTICLE ACCELERATORS↗
DOE OSTI2024

Implementing Superresolution of Nonstationary Tides with Wavelets: An Introduction to CWT_Multi

Abstract Tides are often nonstationary due to nonastronomical influences. Investigating variable tidal properties implies a trade-off between separating adjacent frequencies (using long analysis windows) and resolving their time variations (short analysis windows). Previous continuous wavelet transform (CWT) tidal methods resolved tidal species. Here, we present CWT_Multi, a MATLAB code that 1) uses CWT linearity (via the “response coefficient method”) to implement superresolution, i.e., resolving tidal constituents beyond the Rayleigh criterion; 2) provides a Munk–Hasselmann constituent selection criterion appropriate for superresolution; and 3) introduces an objective, time-variable form of inference (“dynamic inference”) based on time-varying data properties. CWT_Multi resolves tidal species on time scales of days, and multiple constituents per species with fortnightly filters. It outputs astronomical phase lags and admittances, analyzes multiple records, and provides power spectra of the signal(s), residual(s), and reconstruction(s); confidence limits; and signal-to-noise ratios. Artificial data and water levels from the Lower Columbia River Estuary (LCRE) and San Francisco Bay Delta (SFBD) are used to test CWT_Multi and compare it to harmonic analysis programs NS_Tide and UTide. CWT_Multi provides superior reconstruction, detiding, dynamic analysis utility, and time resolution of constituents (but with broader confidence limits). Dynamic inference resolves closely spaced constituents (like K 1 , S 1 , and P 1 ) on fortnightly time scales, quantifying impacts of diel power peaking (with a 24-h period, like S 1 ) on water levels in the LCRE. CWT_Multi also helps quantify the impacts of high flows and a salt barrier closing on tidal properties in the SFBD. On the other hand, CWT_Multi does not excel at prediction, and results depend on analysis details, as for any method applied to nonstationary data. Significance Statement Ocean tides, especially in coastal and estuarine systems, are often nonstationary, in the sense that the mean and standard deviation of tidal properties vary over time, usually in response to some nontidal process. We introduce here a MATLAB code, CWT_Multi, that uses wavelet transforms to resolve both tidal species and constituents on time scales from a few days to months. Our code accommodates multiple scalar time series and has typical tidal analysis features like constituent selection and inference, plus two forms of uncertainty analyses. It is flexible, allowing the user to adapt analysis properties to diverse datasets. CWT_Multi is applicable to many problems involving time-variable tides, including sea level rise, compound flooding, sediment transport, and wetland habitat analyses. Application to vector data is a straightforward extension, but further development of our uncertainty analysis is merited. Because nonstationary tidal analysis is rapidly advancing, we also define the features of a “well-formed” analysis code.

Lobo, Matthew↗
NASA NTRS1986

Image restoration and superresolution as probes of small scale far-IR structure in star forming regions

Far-infrared continuum studies from the Kuiper Airborne Observatory are described that are designed to fully exploit the small-scale spatial information that this facility can provide. This work gives the clearest picture to data on the structure of galactic and extragalactic star forming regions in the far infrared. Work is presently being done with slit scans taken simultaneously at 50 and 100 microns, yielding one-dimensional data. Scans of sources in different directions have been used to get certain information on two dimensional structure. Planned work with linear arrays will allow us to generalize our techniques to two dimensional image restoration. For faint sources, spatial information at the diffraction limit of the telescope is obtained, while for brighter sources, nonlinear deconvolution techniques have allowed us to improve over the diffraction limit by as much as a factor of four. Information on the details of the color temperature distribution is derived as well. This is made possible by the accuracy with which the instrumental point-source profile (PSP) is determined at both wavelengths. While these two PSPs are different, data at different wavelengths can be compared by proper spatial filtering. Considerable effort has been devoted to implementing deconvolution algorithms. Nonlinear deconvolution methods offer the potential of superresolution -- that is, inference of power at spatial frequencies that exceed D lambda. This potential is made possible by the implicit assumption by the algorithm of positivity of the deconvolved data, a universally justifiable constraint for photon processes. We have tested two nonlinear deconvolution algorithms on our data; the Richardson-Lucy (R-L) method and the Maximum Entropy Method (MEM). The limits of image deconvolution techniques for achieving spatial resolution are addressed.

Lester, D. F.↗
NASA NTRS1992

Superresolution via sparsity constraints

The problem of recovering a measure mu supported on a lattice of span Delta is considered under the condition that measurements are only available concerning the Fourier Transform at frequencies of Omega or less. If Omega is much smaller than the Nyquist frequency pi/Delta and the measurements are noisy, then stable recovery of mu is generally impossible. It is shown here that if, in addition, it is known that mu satisfies certain sparsity constraints, then stable recovery is possible. This finding validates practical efforts in spectroscopy, seismic prospecting, and astronomy to provide superresolution by imposing support limitations in reconstruction.

Donoho, David L.↗
DOE OSTI2024

Statistical Treatment of Convolutional Neural Network Superresolution of Inland Surface Wind for Subgrid-Scale Variability Quantification

Abstract Machine learning models have been employed to perform either physics-free data-driven or hybrid dynamical downscaling of climate data. Most of these implementations operate over relatively small downscaling factors because of the challenge of recovering fine-scale information from coarse data. This limits their compatibility with many global climate model outputs, often available between ∼50- and 100-km resolution, to scales of interest such as cloud resolving or urban scales. This study systematically examines the capability of a type of superresolving convolutional neural network (SR-CNNs) to downscale surface wind speed data over land from different coarse resolutions (25-, 48-, and 100-km resolution) to 3 km. For each downscaling factor, we consider three convolutional neural network (CNN) configurations that generate superresolved predictions of fine-scale wind speed, which take between one and three input fields: coarse wind speed, fine-scale topography, and diurnal cycle. In addition to fine-scale wind speeds, probability density function parameters are generated through which sample wind speeds can be generated, accounting for the intrinsic stochasticity of wind speed. For assessing generalization to new data, CNN models are tested on regions with different topography and climate that are unseen during training. The evaluation of superresolved predictions focuses on subgrid-scale variability and the recovery of extremes. Models with coarse wind and fine topography as inputs exhibit the best performance when compared with other model configurations, operating across the same downscaling factor. Our diurnal cycle encoding results in lower out-of-sample generalizability when compared with other input configurations.

17 WIND ENERGY↗
NASA NTRS1999

Nondestructive Superresolution Imaging of Defects and Nonuniformities in Metals, Semiconductors, Dielectrics, Composites, and Plants Using Evanescent Microwaves

We have imaged and mapped material nonuniformities and defects using microwaves generated at the end of a microstripline resonator with 0.4 micrometer lateral spatial resolution at 1 GHz. Here we experimentally examine the effect of microstripline substrate permittivity, the feedline-to-resonator coupling strength, and probe tip geometry on the spatial resolution of the probe. Carbon composites, dielectrics, semiconductors, metals, and botanical samples were scanned for defects, residual stresses, subsurface features, areas of different film thickness, and moisture content. The resulting evanescent microwave probe (EMP) images are discussed. The main objective of this work is to demonstrate the overall capabilities of the EMP imaging technique as well as to discuss various probe parameters that can be used to design EMPs for different applications.

Tabib-Azar, M.↗
NASA NTRS1993

Dynamic testbed laboratory and micro-optics

Objectives of this research are as follows: to understand the three classes of superresolution and their behavior in optical data storage systems; to investigate new and improved components and techniques in the optical system for data detection and servo control; to apply micro-optic components to the optical system to reduce size and weight; to investigate techniques, such as near-field optical probes, for recording data densities beyond that possible with superresolution; and to understand and find solutions for problems associated with dynamic testing, especially those that arise when evaluating blue-sensitive media.

Milster, T.↗
DOE OSTI2024

Report on the AAPM grand challenge on deep generative modeling for learning medical image statistics

Abstract Background The findings of the 2023 AAPM Grand Challenge on Deep Generative Modeling for Learning Medical Image Statistics are reported in this Special Report. Purpose The goal of this challenge was to promote the development of deep generative models for medical imaging and to emphasize the need for their domain‐relevant assessments via the analysis of relevant image statistics. Methods As part of this Grand Challenge, a common training dataset and an evaluation procedure was developed for benchmarking deep generative models for medical image synthesis. To create the training dataset, an established 3D virtual breast phantom was adapted. The resulting dataset comprised about 108 000 images of size 512 512. For the evaluation of submissions to the Challenge, an ensemble of 10 000 DGM‐generated images from each submission was employed. The evaluation procedure consisted of two stages. In the first stage, a preliminary check for memorization and image quality (via the Fréchet Inception Distance [FID]) was performed. Submissions that passed the first stage were then evaluated for the reproducibility of image statistics corresponding to several feature families including texture, morphology, image moments, fractal statistics, and skeleton statistics. A summary measure in this feature space was employed to rank the submissions. Additional analyses of submissions was performed to assess DGM performance specific to individual feature families, the four classes in the training data, and also to identify various artifacts. Results Fifty‐eight submissions from 12 unique users were received for this Challenge. Out of these 12 submissions, 9 submissions passed the first stage of evaluation and were eligible for ranking. The top‐ranked submission employed a conditional latent diffusion model, whereas the joint runners‐up employed a generative adversarial network, followed by another network for image superresolution. In general, we observed that the overall ranking of the top 9 submissions according to our evaluation method (i) did not match the FID‐based ranking, and (ii) differed with respect to individual feature families. Another important finding from our additional analyses was that different DGMs demonstrated similar kinds of artifacts. Conclusions This Grand Challenge highlighted the need for domain‐specific evaluation to further DGM design as well as deployment. It also demonstrated that the specification of a DGM may differ depending on its intended use.

Radiology, Nuclear Medicine & Medical Imaging↗
DOE OSTI2024

Centriole and transition zone structures in photoreceptor cilia revealed by cryo-electron tomography

Primary cilia mediate sensory signaling in multiple organisms and cell types but have structures adapted for specific roles. Structural defects in them lead to devastating diseases known as ciliopathies in humans. Key to their functions are structures at their base: the basal body, the transition zone, the “Y-shaped links,” and the “ciliary necklace.” We have used cryo-electron tomography with subtomogram averaging and conventional transmission electron microscopy to elucidate the structures associated with the basal region of the “connecting cilia” of rod outer segments in mouse retina. The longitudinal variations in microtubule (MT) structures and the lumenal scaffold complexes connecting them have been determined, as well as membrane-associated transition zone structures: Y-shaped links connecting MT to the membrane, and ciliary beads connected to them that protrude from the cell surface and form a necklace-like structure. These results represent a clearer structural scaffold onto which molecules identified by genetics, proteomics, and superresolution fluorescence can be placed in our emerging model of photoreceptor sensory cilia.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Physically Motivated Deep Learning to Superresolve and Cross Calibrate Solar Magnetograms

Abstract Superresolution (SR) aims to increase the resolution of images by recovering detail. Compared to standard interpolation, deep learning-based approaches learn features and their relationships to leverage prior knowledge of what low-resolution patterns look like in higher resolution. Deep neural networks can also perform image cross-calibration by learning the systematic properties of the target images. While SR for natural images aims to create perceptually convincing results, SR of scientific data requires careful quantitative evaluation. In this work, we demonstrate that deep learning can increase the resolution and calibrate solar imagers belonging to different instrumental generations. We convert solar magnetic field images taken by the Michelson Doppler Imager (resolution ∼2″ pixel −1 ; space based) and the Global Oscillation Network Group (resolution ∼2.″5 pixel −1 ; ground based) to the characteristics of the Helioseismic and Magnetic Imager (resolution ∼0.″5 pixel −1 ; space based). We also establish a set of performance measurements to benchmark deep-learning-based SR and calibration for scientific applications.

Muñoz-Jaramillo, Andrés (ORCID:0000000247160840)↗
NASA NTRS1981

Possibility of a new approach to processing space pulse radio signals

A approach to the processing of space pulse radio signals, which makes it possible to achieve a high level of time resolution is proposed. The approach is based on the utilization of the Fourier transform of a radio pulse envelope in the interstellar medium effect. The conditions of applicability of the proposed method and the requirements for recording equipment are substantiated. Also indicated is the possibility of obtaining superresolution with the utilization of the given method of signal processing.

Bocharov, A. A.↗
NASA NTRS1980

Application of modern time series analysis to high stability oscillators

Techniques of modern time series analysis useful for investigating the characteristics of high-stability oscillators and identifying systematic perturbations are discussed with reference to an experiment in which the frequencies of superconducting cavity-stabilized oscillators and hydrogen masers were compared. The techniques examined include transformation to stationarity, autocorrelation and cross-correlation, superresolution, and transfer function determination.

Farrell, B. F.↗
NASA NTRS1986

Far-infrared structure in the nuclear region of M51

Far-infrared (100-micron) scans across the nucleus of M51 are presented. These scans show that much of the far-infrared luminosity of this galaxy comes from a small, sharply bounded region within a 30 arcsec radius (700 pc) of the nucleus. This radius corresponds to the innermost spiral structure, as well as that estimated from dynamical models for the radius of the inner Lindblad resonance. The results of a nonlinear image restoration analysis of the data are presented. This analysis, carried out with two different algorithms, yields information on the spatial structure of star formation near the nucleus that exceeds the diffraction limit of the telescope at this wavelength. This 'superresolution' treatment shows that star formation is enhanced at the outer edge of this region compared to the nucleus, which appears to be relatively devoid of young stars. This ring of active star formation may be produced in the same way as that seen in our own Galaxy at a galactocentric radius of 4-5 kpc and may indicate that the region near the nucleus has been swept clean of material by violent processes.

Lester, D. F.↗
NASA NTRS1999

NASA AMES Remote Operations Center for 2001

There is a Memorandum of Agreement between NASA Ames, JPL, West Virginia University and University of Arizona which led to funding for the MECA microscope and to the establishment of an Ames facility for science analysis of microscopic and other data. The data and analysis will be by agreement of the Mars Environmental Compatibility Assessment (MECA), Robotic Arm Camera (RAC) and other PI's. This facility is intended to complement other analysis efforts with one objective of this facility being to test the latest information technologies in support of actual mission science operations. Additionally, it will be used as a laboratory for the exploration of collaborative science activities. With a goal of enhancing the science return for both Human Exploration and Development of Space (HEDS) and Astrobiology we shall utilize various tools such as superresolution and the Virtual Environment Vehicle Interface (VEVI) virtual reality visualization tools. In this presentation we will describe the current planning for this facility.

Sims, M.↗
NASA NTRS2013

Helicopter Flight Test of a Compact, Real-Time 3-D Flash Lidar for Imaging Hazardous Terrain During Planetary Landing

A second generation, compact, real-time, air-cooled 3-D imaging Flash Lidar sensor system, developed from a number of cutting-edge components from industry and NASA, is lab characterized and helicopter flight tested under the Autonomous Precision Landing and Hazard Detection and Avoidance Technology (ALHAT) project. The ALHAT project is seeking to develop a guidance, navigation, and control (GN&C) and sensing system based on lidar technology capable of enabling safe, precise crewed or robotic landings in challenging terrain on planetary bodies under any ambient lighting conditions. The Flash Lidar incorporates a 3-D imaging video camera based on Indium-Gallium-Arsenide Avalanche Photo Diode and novel micro-electronic technology for a 128 x 128 pixel array operating at a video rate of 20 Hz, a high pulse-energy 1.06 μm Neodymium-doped: Yttrium Aluminum Garnet (Nd:YAG) laser, a remote laser safety termination system, high performance transmitter and receiver optics with one and five degrees field-of-view (FOV), enhanced onboard thermal control, as well as a compact and self-contained suite of support electronics housed in a single box and built around a PC-104 architecture to enable autonomous operations. The Flash Lidar was developed and then characterized at two NASA-Langley Research Center (LaRC) outdoor laser test range facilities both statically and dynamically, integrated with other ALHAT GN&C subsystems from partner organizations, and installed onto a Bell UH-1H Iroquois "Huey" helicopter at LaRC. The integrated system was flight tested at the NASA-Kennedy Space Center (KSC) on simulated lunar approach to a custom hazard field consisting of rocks, craters, hazardous slopes, and safe-sites near the Shuttle Landing Facility runway starting at slant ranges of 750 m. In order to evaluate different methods of achieving hazard detection, the lidar, in conjunction with the ALHAT hazard detection and GN&C system, operates in both a narrow 1deg FOV raster-scanning mode in which successive, gimbaled images of the hazard field are mosaicked together as well as in a wider, 4.85deg FOV staring mode in which digital magnification, via a novel 3-D superresolution technique, is used to effectively achieve the same spatial precision attained with the more narrow FOV optics. The lidar generates calibrated and corrected 3-D range images of the hazard field in real-time and passes them to the ALHAT Hazard Detection System (HDS) which stitches the images together to generate on-the-fly Digital Elevation Maps (DEM's) and identifies hazards and safe-landing sites which the ALHAT GN&C system can then use to guide the host vehicle to a safe landing on the selected site. Results indicate that, for the KSC hazard field, the lidar operational range extends from 100m to 1.35 km for a 30 degree line-of-sight angle and a range precision as low as 8 cm which permits hazards as small as 25 cm to be identified. Based on the Flash Lidar images, the HDS correctly found and reported safe sites in near-real-time during several of the flights. A follow-on field test, planned for 2013, seeks to complete the closing of the GN&C loop for fully-autonomous operations on-board the Morpheus robotic, rocket-powered, free-flyer test bed in which the ALHAT system would scan the KSC hazard field (which was vetted during the present testing) and command the vehicle to landing on one of the selected safe sites.

Roback, VIncent E.↗