Search NASA⌕ Search

SEARCH · Search NASA

Results for “Image processing”

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 991 records · Page 55
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Automated electrosynthesis reaction mining with multimodal large language models (MLLMs)

Leveraging the chemical data available in legacy formats such as publications and patents is a significant challenge for the community. Automated reaction mining offers a promising solution to unleash this knowledge into a learnable digital form and therefore help expedite materials and reaction discovery. However, existing reaction mining toolkits are limited to single input modalities (text or images) and cannot effectively integrate heterogeneous data that is scattered across text, tables, and figures. In this work, we go beyond single input modalities and explore multimodal large language models (MLLMs) for the analysis of diverse data inputs for automated electrosynthesis reaction mining. We compiled a test dataset of 65 articles (MERMES-T24 set) and employed it to benchmark five prominent MLLMs against two critical tasks: (i) reaction diagram parsing and (ii) resolving cross-modality data interdependencies. The frontrunner MLLM achieved ≥96% accuracy in both tasks, with the strategic integration of single-shot visual prompts and image pre-processing techniques. We integrate this capability into a toolkit named MERMES (multimodal reaction mining pipeline for electrosynthesis). Our toolkit functions as an end-to-end MLLM-powered pipeline that integrates article retrieval, information extraction and multimodal analysis for streamlining and automating knowledge extraction. This work lays the groundwork for the increased utilization of MLLMs to accelerate the digitization of chemistry knowledge for data-driven research.

Leong, Shi Xuan↗
DOE OSTI2021

Uranium Bed Design Parameters for Tritium Plants Supporting Fusion Reactors

Uranium hydride is commonly used to store hydrogen or its isotopes in a solid state. The Self-Assaying Tritium Accountancy and Containment Unit for ITER (STACI) is a 5.2-kg bed of depleted uranium (dU) capable of holding up to 33 mol of hydrogen or its isotopes. Here we summarize the data analysis of past experimental campaigns with STACI, with the aim of describing the kinetics and thermodynamics of the hydriding process. Computed tomography imaging was performed on STACI both before and after its experimental campaign, and a high degree of swelling was observed in the dU. Literature on studies in regard to the swelling of large (multikilogram) quantities of uranium hydride for storage applications was not identified during this study. Data from the experimental campaign, as well as data on the formation reaction, are presented. The authors hope to create an analytical model of STACI based on these data.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2020

Small animal PET: a review of what we have done and where we are going

Abstract Small animal research is an essential tool in studying both pharmaceutical biodistribution and disease progression over time. Furthermore, through the rapid development of in vivo imaging technology over the last few decades, small animal imaging (also referred to as preclinical imaging) has become a mainstay for all fields of biologic research and a center point for most preclinical cancer research. Preclinical imaging modalities include optical, MRI and MRS, microCT, small animal PET, ultrasound, and photoacoustic, each with their individual strengths. The strong points of small animal PET are its translatability to the clinic; its quantitative imaging capabilities; its whole-body imaging ability to dynamically trace functional/biochemical processes; its ability to provide useful images with only nano- to pico- molar concentrations of administered compounds; and its ability to study animals serially over time. This review paper gives an overview of the development and evolution of small animal PET imaging. It provides an overview of detector designs; system configurations; multimodality PET imaging systems; image reconstruction and analysis tools; and an overview of research and commercially available small animal PET systems. It concludes with a look toward developing technologies/methodologies that will further enhance the impact of small animal PET imaging on medical research in the future.

Engineering↗
DOE OSTI2024

Optimal 1D Ly α forest power spectrum estimation – III. DESI early data

ABSTRACT The 1D power spectrum P1D of the Ly α forest provides important information about cosmological and astrophysical parameters, including constraints on warm dark matter models, the sum of the masses of the three neutrino species, and the thermal state of the intergalactic medium. We present the first measurement of P1D with the quadratic maximum likelihood estimator (QMLE) from the Dark Energy Spectroscopic Instrument (DESI) survey early data sample. This early sample of 54 600 quasars is already comparable in size to the largest previous studies, and we conduct a thorough investigation of numerous instrumental and analysis systematic errors to evaluate their impact on DESI data with QMLE. We demonstrate the excellent performance of the spectroscopic pipeline noise estimation and the impressive accuracy of the spectrograph resolution matrix with 2D image simulations of raw DESI images that we processed with the DESI spectroscopic pipeline. We also study metal line contamination and noise calibration systematics with quasar spectra on the red side of the Ly α emission line. In a companion paper, we present a similar analysis based on the Fast Fourier Transform estimate of the power spectrum. We conclude with a comparison of these two approaches and discuss the key sources of systematic error that we need to address with the upcoming DESI Year 1 analysis.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2018

Superresolution Diffuse Optical Imaging by Localization of Fluorescence

The multiple scattering of light presents major challenges in realizing useful in vivo imaging at tissue depths of more than about one millimeter, where many answers to health questions lie. Visible through near-infrared photons can be readily and safely detected through centimeters of tissue; however, limited information is available for image formation. One strategy for obtaining images is to model the photon transport and a simple incoherent model is the diffusion equation approximation to the Boltzmann transport equation. Such an approach provides a prediction of the mean intensity of heavily scattered light and hence provides a forward model for optimization-based computational imaging. While diffuse optical imaging methods have received substantial attention, they remain restricted in terms of resolution because of the loss of high-spatial-frequency information that is associated with the multiple scattering of photons. Consequently, only relatively large inhomogeneities, such as tumors or organs in small animals, can be effectively resolved. We introduce a superresolution imaging approach based on point localization in a diffusion framework that enables over two orders of magnitude improvement in the spatial resolution of diffuse optical imaging. The method is demonstrated experimentally by localizing a fluorescent inhomogeneity in a highly scattering slab and characterizing the localization uncertainty. The approach allows imaging through centimeters of tissue with a resolution of tens of microns, thereby enabling cells or cell clusters to be resolved. More generally, this high-resolution imaging approach could be applied with any physical transport or wave model and hence to a broad class of physical problems. Finally, paired with a suitable optical contrast mechanism, as can be realized with targeted fluorescent molecules or genetically modified animals, superresolution diffuse imaging should open alternative dimensions for in vivo applications.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2024

Bragg Spot Finder (BSF): a new machine-learning-aided approach to deal with spot finding for rapidly filtering diffraction pattern images

Macromolecular crystallography contributes significantly to understanding diseases and, more importantly, how to treat them by providing atomic resolution 3D structures of proteins. This is achieved by collecting X-ray diffraction images of protein crystals from important biological pathways. Spotfinders are used to detect the presence of crystals with usable data, and the spots from such crystals are the primary data used to solve the relevant structures. Having fast and accurate spot finding is essential, but recent advances in synchrotron beamlines used to generate X-ray diffraction images have brought us to the limits of what the best existing spotfinders can do. This bottleneck must be removed so spotfinder software can keep pace with the X-ray beamline hardware improvements and be able to see the weak or diffuse spots required to solve the most challenging problems encountered when working with diffraction images. In this paper, we first present Bragg Spot Detection (BSD), a large benchmark Bragg spot image dataset that contains 304 images with more than 66 000 spots. We then discuss the open source extensible U-Net-based spotfinder Bragg Spot Finder (BSF), with image pre-processing, a U-Net segmentation backbone, and post-processing that includes artifact removal and watershed segmentation. Finally, we perform experiments on the BSD benchmark and obtain results that are (in terms of accuracy) comparable to or better than those obtained with two popular spotfinder software packages ( Dozor and DIALS ), demonstrating that this is an appropriate framework to support future extensions and improvements.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Nanoscale chromatin imaging and analysis platform bridges 4D chromatin organization with molecular function

Extending across multiple length scales, dynamic chromatin structure is linked to transcription through the regulation of genome organization. However, no individual technique can fully elucidate this structure and its relation to molecular function at all length and time scales at both a single-cell level and a population level. Here, we present a multitechnique nanoscale chromatin imaging and analysis (nano-ChIA) platform that consolidates electron tomography of the primary chromatin fiber, optical super-resolution imaging of transcription processes, and label-free nano-sensing of chromatin packing and its dynamics in live cells. Using nano-ChIA, we observed that chromatin is localized into spatially separable packing domains, with an average diameter of around 200 nanometers, sub-megabase genomic size, and an internal fractal structure. The chromatin packing behavior of these domains exhibits a complex bidirectional relationship with active gene transcription. Furthermore, we found that properties of PDs are correlated among progenitor and progeny cells across cell division.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Meta-optic accelerators for object classifiers

Rapid advances in deep learning have led to paradigm shifts in a number of fields, from medical image analysis to autonomous systems. These advances, however, have resulted in digital neural networks with large computational requirements, resulting in high energy consumption and limitations in real-time decision-making when computation resources are limited. Here, we demonstrate a meta-optic–based neural network accelerator that can off-load computationally expensive convolution operations into high-speed and low-power optics. In this architecture, metasurfaces enable both spatial multiplexing and additional information channels, such as polarization, in object classification. End-to-end design is used to co-optimize the optical and digital systems, resulting in a robust classifier that achieves 93.1% accurate classification of handwriting digits and 93.8% accuracy in classifying both the digit and its polarization state. This approach could enable compact, high-speed, and low-power image and information processing systems for a wide range of applications in machine vision and artificial intelligence.

42 ENGINEERING↗
DOE OSTI2022

SkX_NN: Neural Network for skyrmion identification in Lorentz TEM

A neural network for segmenting and processing Lorentz TEM images of skyrmion lattices. This software based on U-Net architecture is capable of identifying skyrmions in out-of-focus Lorentz TEM images based on training data. This isspecifically useful for analyzing large area data from in-situ experiments.

PHATAK, CHARUDATTA↗
DOE OSTI2024

recon3d

SAND2025-00533O recon3d is a software tool that provides automated 3D reconstruction and meshing capabilities. It processes labeled 3D image data from various sources, starting from image stacks, and calculates 3D feature distributions like size, shape, and location. The software also has tools for downscaling rectilinear grid data and creating tetrahedral meshes directly from image data. recon3d can be used by novice users via the command line with a properly formatted configuration file. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Emery, John↗
DOE OSTI2025

GRinding Automated Classification Engine

This work is an ML-driven framework for automated surface analysis of microscopy images. We create a training dataset by imaging stainless steel samples to benchmark four developed deep neural network architectures. These models, based on a YOLOv8n-cls backend, integrate image features and process metadata using various fusion methods to distinguish between acceptable and unacceptable surface finishes. This code is associated with publication "Classifying Alloy Surface Preparation Quality with Metadata-Infused Machine Learning for Rapid Alloy Discovery" for project APEX LDRD-ER (25-ERD-039)

Gongora, AldairE [Lawrence Livermore National Labo↗
DOE OSTI2021

Development of a Random-Forest Cloud-Regime Classification Model Based on Surface Radiation and Cloud Products

Various methods have been developed to characterize cloud type, otherwise referred to as cloud regime. These include manual sky observations, combining radiative and cloud vertical properties observed from satellite, surface-based remote sensing, and digital processing of sky imagers. While each method has inherent advantages and disadvantages, none of these cloud-typing methods actually includes measurements of surface shortwave or longwave radiative fluxes. Here, a method that relies upon detailed, surface-based radiation and cloud measurements and derived data products to train a random-forest machine-learning cloud classification model is introduced. Measurements from five years of data from the ARM Southern Great Plains site were compiled to train and independently evaluate the model classification performance. A cloud-type accuracy of approximately 80% using the random-forest classifier reveals that the model is well suited to predict climatological cloud properties. Furthermore, an analysis of the cloud-type misclassifications is performed. While physical cloud types may be misreported, the shortwave radiative signatures are similar between misclassified cloud types. From this, we assert that the cloud-regime model has the capacity to successfully differentiate clouds with comparable cloud–radiative interactions. Therefore, we conclude that the model can provide useful cloud-property information for fundamental cloud studies, inform renewable energy studies, and be a tool for numerical model evaluation and parameterization improvement, among many other applications.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

UAV remote sensing imagery - Miscanthus trials 2020 - Energy Farm - UIUC

Aerial imagery utilized as input in the manuscript "Deep convolutional neural networks exploit high spatial and temporal resolution aerial imagery to predict key traits in miscanthus" . Data was collected over M. Sacchariflorus and Sinensis breeding trials at the Energy Farm, UIUC in 2020. Flights were performed using a DJI M600 mounted with a Micasense Rededge multispectral sensor at 20 m altitude around solar noon. Imagery is available as tif file by field trial and date (10). The post-processing of raw images into orthophoto was performed in Agisoft Metashape software. Each crop surface model and multispectral orthophoto was stacked into an unique raster stack by date and uploaded here. Each raster stack includes 6 layers in the following order: Layer 1 = crop surface model, Layer 2 = Blue, Layer 3 = Green, Layer 4 = Red, Layer 5 = Rededge, and Layer 6 = NIR multispectral bands. Msa raster stacks were resampled to 1.67 cm spatial resolution and Msi raster stacks were resampled to 1.41 cm spatial resolution to ease their integration into further analysis. 'MMDDYYYY' is the date of data collection, 'MSA' is M. Sacchariflorus trial, 'MSI' is Miscanthus Sinensis trial, 'CSM' is crop surface model layer, and 'MULTSP' are the five multispectral bands.

bioenergy↗
DOE OSTI2026

Adaptive nonlinear wavefront shaping in layered TMDs

Two-dimensional transition metal dichalcogenides are promising candidates for nonlinear photonics applications, offering strong nonlinear susceptibility and compatibility with integrated platforms. Although considerable effort has been devoted to manipulating second harmonic generation in these materials, the dynamic control of nonlinear wavefronts remains largely unexplored. Metasurfaces have enabled significant progress in nonlinear beam engineering, yet their practical implementation is limited by low conversion efficiencies and restricted design flexibility. In this work, we employ feedback-based wavefront shaping using spatial light modulators to enhance SHG from pyramid-like WS 2 multilayer structures by over three orders of magnitude in selected spatial regions. This approach allows us to program nonlinear holograms and dynamically shape the SHG signal through phase-only modulation, opening new possibilities for nonlinear imaging, optical information processing, and data communication.

Mavian, Alex [Rensselaer Polytechnic Inst., Troy, ↗
DOE OSTI2021

Giant terahertz polarization rotation in ultrathin films of aligned carbon nanotubes

For easy manipulation of polarization states of light for applications in communications, imaging, and information processing, an efficient mechanism is desired for rotating light polarization with a minimum interaction length. Here, we report giant polarization rotations for terahertz (THz) electromagnetic waves in ultrathin ( ∼ <#comment/> 45 n m ), high-density films of aligned carbon nanotubes. We observed polarization rotations of up to ∼ <#comment/> 20 ∘ <#comment/> and ∼ <#comment/> 110 ∘ <#comment/> for transmitted and reflected THz pulses, respectively. The amount of polarization rotation was a sensitive function of the angle between the incident THz polarization and the nanotube alignment direction, exhibiting a “magic” angle at which the total rotation through transmission and reflection becomes exactly 90°. Our model quantitatively explains these giant rotations as a result of extremely anisotropic optical constants, demonstrating that aligned carbon nanotubes promise ultrathin, broadband, and tunable THz polarization devices.

Baydin, Andrey (ORCID:0000000225673506)↗
DOE OSTI2018

An effective “three-in-one” screening assay for testing drug and nanoparticle toxicity in human endothelial cells

Evaluating nanoparticle (NP) toxicity in human cell systems is a fundamental requirement for future NP biomedical applications. In this study, we have designed a screening assay for assessing different types of cell death induced by NPs in human umbilical vein endothelial cell (HUVEC) culture. This assay consists of WST-8, LDH and Hoechst 33342 staining, all performed in one well, which enables an evaluation of cell viability, necrosis and apoptosis, respectively, in the same cell sample. The 96-well format and automated processing of fluorescent images enhances the assay rapidity and reproducibility. After testing the assay functionality with agents that induced different types of cell death, we investigated the endothelial toxicity of superparamagnetic iron oxide nanoparticles (SPIONs, 8 nm), silica nanoparticles (SiNPs, 7–14 nm) and carboxylated multiwall carbon nanotubes (CNTCOOHs, 60 nm). Our results indicated that all the tested NP types induced decreases in cell viability after 24 hours at a concentration of 100 μg/ml. SPIONs caused the lowest toxicity in HUVECs. By contrast, SiNPs induced pronounced necrosis and apoptosis. A time course experiment showed the gradual toxic effect of all the tested NPs. CNTCOOHs inhibited tetrazolium derivatives at 100 μg/ml, causing false negative results from the WST-8 and LDH assay. In summary, our data demonstrate that the presented “three-in-one” screening assay is capable of evaluating NP toxicity effectively and reliably. Due to its simultaneous utilization of two different methods to assess cell viability, this assay is also capable of revealing, if NPs interfere with tetrazolium salts.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2017

Imaging of Nuclear Weapon Trainers

The Configurable Muon Tracker (CMT) is an adaptation of the existing drift tube detector commercially available from Decision Sciences International Corporation (DSIC). NSTec engineered the CMT around commercially available drift tube assemblies to make a detector that is more versatile than previous drift tube assemblies. The CMT became operational in February 2013. Traditionally, cosmic-ray muon trackers rely on near-vertical trajectory muons for imaging. Since there are scenarios where imaging using vertical trajectory muons is not practical, NSTec designed the CMT specifically for quick configurability to track muons from any trajectory. The CMT was originally designed to be changed from vertical imaging mode to horizontal imaging mode in a few hours with access to a crane or other lifting equipment. In FY14, locations for imaging weapon trainers and SNM were identified and it was determined that lifting equipment would not typically be available in experimental areas. The CMT was further modified and a portable lifting system was developed to allow reconfiguration of the CMT without access to lifting equipment at the facility. This system was first deployed at Los Alamos National Laboratory’s W-division, where several trainers were imaged in both horizontal and vertical modes. Real-time images have been compared in both modes showing that imaging can be done in both modes with the expected longer integration time for horizontal mode. Further imaging and post processing of the data is expected to continue into early FY15.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗