Search NASA⌕ Search

SEARCH · Search NASA

Results for “Reconstruction algorithms”

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 1,315 records · Page 73

Tomography by iterative convolution - Empirical study and application to interferometry

An algorithm for computer tomography has been developed that is applicable to reconstruction from data having incomplete projections because an opaque object blocks some of the probing radiation as it passes through the object field. The algorithm is based on iteration between the object domain and the projection (Radon transform) domain. Reconstructions are computed during each iteration by the well-known convolution method. Although it is demonstrated that this algorithm does not converge, an empirically justified criterion for terminating the iteration when the most accurate estimate has been computed is presented. The algorithm has been studied by using it to reconstruct several different object fields with several different opaque regions. It also has been used to reconstruct aerodynamic density fields from interferometric data recorded in wind tunnel tests.

Vest, C. M.↗

Real-Time Artificial Intelligence for Particle Reconstruction and Higgs Physics

With the discovery of the Higgs boson at the CERN LHC, the world's highest-energy particle accelerator complex, scientists have acquired an important tool to study the fundamental building blocks of the universe. Precision measurements of Higgs bosons produced with large momentum allow for unique insights into the structure of the interactions of the Higgs boson with other particles that may shed light on physics beyond the standard model. While experimentally challenging, exploring such interactions with novel artificial intelligence (AI) methods can advance our understanding of the Higgs sector, including the Higgs boson's self-interaction. Moreover, the LHC is undergoing a major upgrade to further increase its particle collision rate and thereby operate for an additional decade. The experimental detectors at the upgraded facility must process at least a factor of ten more data at rates of hundreds of terabytes per second all under challenging conditions. New AI techniques are required to reconstruct and select, or trigger on, the most physics-sensitive events in real-time to handle the resulting avalanche of data. The proposed research will achieve the goals of the LHC program at the CMS experiment by developing a sub-microsecond event reconstruction system using real-time AI algorithms that employ field-programmable gate array technologies. By harnessing sophisticated AI techniques, this research focuses on measuring the production of Higgs bosons at large momentum while enhancing particle reconstruction methods in the trigger and beyond. Overall, the proposed research has broader implications for the use of AI in resource-constrained, low-latency embedded applications across all fields of science.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Downscaling-Based Segmentation for Unresolved Images of Highly Heterogeneous Granular Porous Samples

Numerical simulations of pore-scale flow and transport in natural sediments require the knowledge of pore-space topology. Limited resolution of X-ray tomography is often insufficient to fully characterize pore-space structure within fine-grained regions. Single and multilevel threshold-based segmentation approaches are customarily employed to identify solid, pore and porous-solid regions by means of grey intensity thresholds. While the choice of cutoff thresholds is often arbitrary, it dramatically affects the effective properties and the dynamical response of the reconstructed porous structure. As such, we propose an algorithm of downscaling, i.e., the process of increasing image resolution, followed by segmentation, i.e., the identification of different phases, to reconstruct the unresolved pore-space from XCT images of natural geological porous media. The method, applicable to moderately unresolved, chemically homogeneous granular media, is based on a map between local pixel porosity and pore size that does not rely on the definition of arbitrary thresholds and it allows to generate a high-resolution binary image of the porous medium from poorly resolved grey-scale images. First, we validate the method on synthetic unresolved images and compare their known pore-space distribution with the extracted one. Then, we consider a synthetic porous medium and compare the pore size distribution, conductivity, and tortuosity between the original and the reconstructed structures. Finally, we apply the method to extract the pore-space distribution from unresolved XCT images of two natural sediment columns and use it (i) to parametrize a capillary-bundle model and (ii) to estimate the hydraulic conductivity by matching breakthrough behavior of passive solute transport.

58 GEOSCIENCES↗

Sci—Thur PM: Imaging — 05: Calibration of a SPECT/CT camera for quantitative SPECT with {sup 99m}Tc

While quantitation is the norm in PET, it is not widely available yet in SPECT. This work's aim was to calibrate a commercially available SPECT/CT system to perform quantitative SPECT. Counting sensitivity, dead-time (DT) constant and partial volume effect (PVE) of the system were assessed. A dual-head Siemens SymbiaT6 SPECT/CT camera equipped with low energy high-resolution collimators was studied. {sup 99m}Tc was the radioisotope of interest because of its wide usage in nuclear medicine. First, point source acquisitions were performed (activity: 30–990MBq). Further acquisitions were then performed with a uniform Jaszczak phantom filled with water at high activity (25–5000MBq). PVE was studied using 6 hot spheres (diameters: 9.9–31.2 mm) filled with {sup 99m}Tc (2.8MBq/cc) in the Jaszczak phantom, which was: (1) empty, (2) water-filled and (3) water-filled with low activity (0.1MBq/cc). The data was reconstructed with the Siemens's Flash3D iterative algorithm with 4 subsets and 8 iterations, attenuation-correction (AC) and scatter-correction (SC). DT modelling was based on the total spectrum counting rate. Sensitivity was assessed using AC-SC reconstructed SPECT data. Sensitivity and DT for the sources were 99.51±1.46cps/MBq and 0.60±0.04µs. For the phantom, sensitivity and DT were 109.9±2.3cps/MBq and 0.62±0.13µs. The recovery-coefficient varied from 5% for the 9.9mm, to 80% for the 31.2mm spheres. With our calibration methods, both sensitivity and DT constant of the SPECT camera had little dependence on the object geometry and attenuation. For small objects of known size, recovery-coefficient can be applied to correct PVE. Clinical quantitative SPECT appears to be possible and has many potential applications.

07 ISOTOPE AND RADIATION SOURCES↗

Direct Numerical Simulation of Acoustic Waves Interacting with a Shock Wave in a Quasi-1D Convergent-Divergent Nozzle Using an Unstructured Finite Volume Algorithm

Numerical simulation of a very small amplitude acoustic wave interacting with a shock wave in a quasi-1D convergent-divergent nozzle is performed using an unstructured finite volume algorithm with a piece-wise linear, least square reconstruction, Roe flux difference splitting, and second-order MacCormack time marching. First, the spatial accuracy of the algorithm is evaluated for steady flows with and without the normal shock by running the simulation with a sequence of successively finer meshes. Then the accuracy of the Roe flux difference splitting near the sonic transition point is examined for different reconstruction schemes. Finally, the unsteady numerical solutions with the acoustic perturbation are presented and compared with linear theory results.

Bui, Trong T.↗

Energy functions for regularization algorithms

Regularization techniques are widely used for inverse problem solving in computer vision such as surface reconstruction, edge detection, or optical flow estimation. Energy functions used for regularization algorithms measure how smooth a curve or surface is, and to render acceptable solutions these energies must verify certain properties such as invariance with Euclidean transformations or invariance with parameterization. The notion of smoothness energy is extended here to the notion of a differential stabilizer, and it is shown that to void the systematic underestimation of undercurvature for planar curve fitting, it is necessary that circles be the curves of maximum smoothness. A set of stabilizers is proposed that meet this condition as well as invariance with rotation and parameterization.

Delingette, H.↗

Water Management In PEM Fuel Cell -“ A Lattice-Boltzmann Modeling Approach

In Proton Exchange Membrane Fuel Cells (PEMFCs), water management and the effective transport of water through the gas-diffusion-layer (GDL) are key issues for improved performance at high power density and for durability during freeze-thaw cycles. The diffusion layer is a thin (~150-350{micro}m), porous material typically composed of a web of carbon fibers and particles, and is usually coated with hydrophobic Teflon to remove the excess water through capillary action. In-situ diagnostics of water movement and gas-reactant transport through this thin opaque substrate is challenging. Numerical analyses are typically based on simplified assumptions, such as Darcy's Law and Leverett functions for the capillary pressure. The objective of this work is to develop a high fidelity CFD modeling and analysis tool to capture the details of multiphase transport through the porous GDL. The tool can be utilized to evaluate GDL material design concepts and optimize systems based on the interactions between cell design, materials, and operating conditions. The flow modeling is based on the Lattice Boltzmann Method (LBM). LBM is a powerful modeling tool to simulate multiphase flows. Its strength is in its kinetic theory based foundation, which provides a fundamental basis for incorporating intermolecular forces that lead to liquid-gas phase separation and capillary effects without resorting to expensive or ad-hoc interface reconstruction schemes. At the heart of the solution algorithm is a discrete form of the well-known Boltzmann Transport Equation (BTE) for molecular distribution, tailored to recover the continuum Navier-Stokes flow. The solution advances by a streaming and collision type algorithm, mimicking actual molecular physics, which makes it suitable for porous media involving complex boundaries. We developed a numerical scheme to reconstruct various porous GDL microstructures including Teflon loading. Single and multiphase LBM models are implemented to compute permeability. Predicted values are in good agreement with measured data. The present modeling approach resolves the GDL microstructures and captures the influence of fiber orientation on permeability and the influence of Teflon loading on the development of preferential flow paths through the GDL. These observations can potentially guide the development of novel GDL materials designed for efficient removal of water.

30 DIRECT ENERGY CONVERSION↗

Performance of the Moving Voxel Image Reconstruction (MVIR) Method in the Fixed Site Detection System (FSDS) Prototype

We have developed a dynamic image reconstruction method called MVIR (Moving Voxel Image Reconstruction) for lane detection in multilane portal monitor systems. MVIR was evaluated for use in the Fixed Site Detection System, a prototype three-lane portal monitor system for EZ-pass toll plazas. As a baseline, we compared MVIR with a static image reconstruction method in analyzing the same real and simulated data sets. Performance was judged by the distributions of image intensities for source and no-source vehicles over many trials as a function of source strength. We found that MVIR produced significantly better results in all cases. The performance difference was greatest at low count rates, where source/no-source distributions were well separated with the MVIR method, allowing reliable source vehicle identification with a low probability of false positive identifications. Static reconstruction of the same data produced overlapping distributions that made source vehicle identification unreliable. The performance of the static method was acceptable at high count rates. Both algorithms reliably identified two strong sources passing through at nearly the same time.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

A Parallel, Finite-Volume Algorithm for Large-Eddy Simulation of Turbulent Flows

A parallel, finite-volume algorithm has been developed for large-eddy simulation (LES) of compressible turbulent flows. This algorithm includes piecewise linear least-square reconstruction, trilinear finite-element interpolation, Roe flux-difference splitting, and second-order MacCormack time marching. Parallel implementation is done using the message-passing programming model. In this paper, the numerical algorithm is described. To validate the numerical method for turbulence simulation, LES of fully developed turbulent flow in a square duct is performed for a Reynolds number of 320 based on the average friction velocity and the hydraulic diameter of the duct. Direct numerical simulation (DNS) results are available for this test case, and the accuracy of this algorithm for turbulence simulations can be ascertained by comparing the LES solutions with the DNS results. The effects of grid resolution, upwind numerical dissipation, and subgrid-scale dissipation on the accuracy of the LES are examined. Comparison with DNS results shows that the standard Roe flux-difference splitting dissipation adversely affects the accuracy of the turbulence simulation. For accurate turbulence simulations, only 3-5 percent of the standard Roe flux-difference splitting dissipation is needed.

Bui, Trong T.↗

Identification of Shielding Material Configurations Using NMIS Imaging

The Nuclear Materials Identification System (NMIS) uses fast neutron tomographic imaging to nonintrusively examine the interior structure of shielded objects. The pixel values in such images represent the attenuation coefficients of the time and directionally tagged fast neutrons from a deuterium-tritium (DT) neutron generator. The reconstruction techniques use either a filtered back projection or a maximum likelihood expectation maximization algorithm. As a first test of the capabilities of these reconstruction techniques to correctly identify individual parts inside of an object, fast neutron imaging was used to identify the regions of shielding surrounding a depleted uranium casting from a library of possible parts. The shielding consisted of multiple regions of common materials such as steel, lead, aluminum, and polyethylene. First, the full object was imaged, and then each of the individual parts was imaged. Several additional parts that were not present in the original object were also imaged to form a library. The individual parts were compared to the full object, and the correct ones were identified using three different methods. These methods included a visual match, an iterative fit of each part, and a mathematical test comparing the sum of squared errors. The successful results demonstrate an initial application of matching. This suggests that it should be possible to implement more sophisticated matching techniques using automated pixel-by-pixel comparison methods in the future.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Reconstructing magnetic deflections from sets of proton images using differential evolution

Proton imaging is a powerful technique for imaging electromagnetic fields within an experimental volume, in which spatial variations in proton fluence are a result of deflections to proton trajectories due to interaction with the fields. When deflections are large, proton trajectories can overlap, and this nonlinearity creates regions of greatly increased proton fluence on the image, known as caustics. The formation of caustics has been a persistent barrier to reconstructing the underlying fields from proton images. We have developed a new method for reconstructing the path-integrated magnetic fields, which begins to address the problem posed by caustics. Our method uses multiple proton images of the same object, each image at a different energy, to fill in the information gaps and provide some uniqueness when reconstructing caustic features. We use a differential evolution algorithm to iteratively estimate the underlying deflection function, which accurately reproduces the observed proton fluence at multiple proton energies simultaneously. We test this reconstruction method using synthetic proton images generated for three different, cylindrically symmetric field geometries at various field amplitudes and levels of proton statistics and present reconstruction results from a set of experimental images. Here, the method we propose requires no assumption of deflection linearity and can reliably solve for fields underlying linear, nonlinear, and caustic proton image features for the selected geometries and is shown to be fairly robust to noise in the input proton intensity.

47 OTHER INSTRUMENTATION↗

An experimental comparison of various methods of nearfield acoustic holography

An experimental comparison of four different methods of nearfield acoustic holography (NAH) is presented in this study for planar acoustic sources. The four NAH methods considered in this study are based on: (1) spatial Fourier transform, (2) equivalent sources model, (3) boundary element methods and (4) statistically optimized NAH. Two dimensional measurements were obtained at different distances in front of a tonal sound source and the NAH methods were used to reconstruct the sound field at the source surface. Reconstructed particle velocity and acoustic pressure fields presented in this study showed that the equivalent sources model based algorithm along with Tikhonov regularization provided the best localization of the sources. Reconstruction errors were found to be smaller for the equivalent sources model based algorithm and the statistically optimized NAH algorithm. Effect of hologram distance on the performance of various algorithms is discussed in detail. The study also compares the computational time required by each algorithm to complete the comparison. Four different regularization parameter choice methods were compared. The L-curve method provided more accurate reconstructions than the generalized cross validation and the Morozov discrepancy principle. Finally, the performance of fixed parameter regularization was comparable to that of the L-curve method.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Design of a multi-detector, single line-of-sight, time-of-flight system to measure time-resolved neutron energy spectra

In the dynamic environment of burning, thermonuclear deuterium–tritium plasmas, diagnosing the time-resolved neutron energy spectrum is of critical importance. Strategies exist for this diagnosis in magnetic confinement fusion plasmas, which presently have a lifetime of ~10 12 longer than inertial confinement fusion (ICF) plasmas. In this work, we present a novel concept for a simple, precise, and scale-able diagnostic to measure time-resolved neutron spectra in ICF plasmas. The concept leverages general tomographic reconstruction techniques adapted to time-of-flight parameter space, and then employs an updated Monte Carlo algorithm and National Ignition Facility-relevant constraints to reconstruct the time-evolving neutron energy spectrum. Reconstructed spectra of the primary 14.028 MeV n DT peak are in good agreement with the exact synthetic spectra. The technique is also used to reconstruct the time-evolving downscattered spectrum, although the present implementation shows significantly more error.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Testing light-traces-mass in Hubble Frontier Fields Cluster MACS-J0416.1-2403

Here, we reconstruct the projected mass distribution of a massive merging Hubble Frontier Fields cluster MACSJ0416 using the genetic algorithm based free-form technique called Grale. The reconstructions are constrained by 149 lensed images identified by Jauzac et al. using HFF data. No information about cluster galaxies or light is used, which makes our reconstruction unique in this regard. Using visual inspection of the maps, as well as galaxy-mass correlation functions we conclude that overall light does follow mass. Furthermore, the fact that brighter galaxies are more strongly clustered with mass is an important confirmation of the standard biasing scenario in galaxy clusters. On the smallest scales, approximately less than a few arcseconds, the resolution afforded by 149 images is still not sufficient to confirm or rule out galaxy-mass offsets of the kind observed in ACO 3827. We also compare the mass maps of MACSJ0416 obtained by three different groups: Grale, and two parametric Lenstool reconstructions from the CATS and Sharon/Johnson teams. Overall, the three agree well; one interesting discrepancy between Grale and Lenstool galaxy-mass correlation functions occurs on scales of tens of kpc and may suggest that cluster galaxies are more biased tracers of mass than parametric methods generally assume.

79 ASTRONOMY AND ASTROPHYSICS↗

Dual X-ray computed tomography-aided classification of melt pool boundaries and flaws in crept additively manufactured parts

In metal additive manufacturing (AM), understanding the process-structure-performance relationships requires a combination of multi-scale characterization techniques that allows for the measurement of the melt pool shape and boundary and classifying various defects and flaws in the AM parts. Such approaches can be destructive, only 2D in nature, or have a small field of view and can be complex to co-register and analyze. Here, in this work, we present a non-destructive 3D inspection technique that employs dual-energy X-ray computed tomography (XCT) along with a model-based iterative reconstruction (MBIR) and a new segmentation algorithm. The proposed approach and algorithm are not only capable of classifying and quantifying flaws such as pores, cracks, and inclusions, but they also allow for the extraction of microstructural features such as melt pool boundaries (MPB) and melt pool regions (MPR), that can help understand process-structure-performance relationships for alloys under study. As an exemplar application, we employed the method for characterization of an additively manufactured aluminum alloy crept under tensile stress at 300 °C for 1064 h. Our results demonstrate high quality segmentation and classification of various flaws and MPB and MPR, for the first time, using 3D X-ray CT inspection. The delineated MPB and MPR in the crept samples reveal the preferential growth paths of cracks that formed during creep deformation. The technique was used for successfully quantifying the characteristics (number of defects, their density, volume fraction, etc.) of the manufacturing-induced pores and creep-induced cracks, which is necessary to better understand the creep failure mechanisms of the material.

36 MATERIALS SCIENCE↗

Dynamic X-ray diffraction sampling for protein crystal positioning

A sparse supervised learning approach for dynamic sampling (SLADS) is described for dose reduction in diffraction-based protein crystal positioning. Crystal centering is typically a prerequisite for macromolecular diffraction at synchrotron facilities, with X-ray diffraction mapping growing in popularity as a mechanism for localization. In X-ray raster scanning, diffraction is used to identify the crystal positions based on the detection of Bragg-like peaks in the scattering patterns; however, this additional X-ray exposure may result in detectable damage to the crystal prior to data collection. Dynamic sampling, in which preceding measurements inform the next most information-rich location to probe for image reconstruction, significantly reduced the X-ray dose experienced by protein crystals during positioning by diffraction raster scanning. The SLADS algorithm implemented herein is designed for single-pixel measurements and can select a new location to measure. In each step of SLADS, the algorithm selects the pixel, which, when measured, maximizes the expected reduction in distortion given previous measurements. Ground-truth diffraction data were obtained for a 5 µm-diameter beam and SLADS reconstructed the image sampling 31% of the total volume and only 9% of the interior of the crystal greatly reducing the X-ray dosage on the crystal. Furthermore, by using in situ two-photon-excited fluorescence microscopy measurements as a surrogate for diffraction imaging with a 1 µm-diameter beam, the SLADS algorithm enabled image reconstruction from a 7% sampling of the total volume and 12% sampling of the interior of the crystal. When implemented into the beamline at Argonne National Laboratory, without ground-truth images, an acceptable reconstruction was obtained with 3% of the image sampled and approximately 5% of the crystal. The incorporation of SLADS into X-ray diffraction acquisitions has the potential to significantly minimize the impact of X-ray exposure on the crystal by limiting the dose and area exposed for image reconstruction and crystal positioning using data collection hardware present in most macromolecular crystallography end-stations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Lossy compression of statistical data using quantum annealer

Abstract We present a new lossy compression algorithm for statistical floating-point data through a representation learning with binary variables. The algorithm finds a set of basis vectors and their binary coefficients that precisely reconstruct the original data. The optimization for the basis vectors is performed classically, while binary coefficients are retrieved through both simulated and quantum annealing for comparison. A bias correction procedure is also presented to estimate and eliminate the error and bias introduced from the inexact reconstruction of the lossy compression for statistical data analyses. The compression algorithm is demonstrated on two different datasets of lattice quantum chromodynamics simulations. The results obtained using simulated annealing show 3–3.5 times better compression performance than the algorithm based on neural-network autoencoder. Calculations using quantum annealing also show promising results, but performance is limited by the integrated control error of the quantum processing unit, which yields large uncertainties in the biases and coupling parameters. Hardware comparison is further studied between the previous generation D-Wave 2000Q and the current D-Wave Advantage system. Our study shows that the Advantage system is more likely to obtain low-energy solutions for the problems than the 2000Q.

97 MATHEMATICS AND COMPUTING↗