SEARCH · Engineering Papers

Results for “Image representation”

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.

SourceResearch connectionFrom yearThrough year

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

At least 217 records · Page 12

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Data Files for Runoff Evaluation in an Earth System Land Model for Permafrost Regions

Modeling of hydrological runoff is essential for accurately capturing spatiotemporal feedbacks within the land–atmosphere system, particularly in sensitive regions such as permafrost landscapes. However, substantial uncertainties persist in the terrestrial runoff parameterization schemes used in Earth system and land surface models. This is particularly true in permafrost regions, where landscape heterogeneity is high and reliable observational data are scarce.This data set includes all files that were produced and applied in the paper Runoff Evaluation in an Earth System Land Model for Permafrost Regions [Xiang et al. in review]. The paper is in review as of July 1 2025 in Geoscientific Model Development (GMD). In this study, we evaluate the performance of runoff parameterization schemes in the Energy Exascale Earth System Model (E3SM) land model (ELM). Our proposed framework leverages simulation results from the Advanced Terrestrial Simulator (ATS), which is a physics-rich integrated surface/subsurface hydrologic model that has been successfully evaluated previously in Arctic tundra regions. We used ATS to simulate runoff from 22 representative hillslopes in the Sagavanirktok River basin, located on the North Slope of Alaska, then compared the output with ELM’s parameterized representation of total runoff. This dataset contains 2 figure image files (*.png, *jpg) that describe the study site and methods, as well as folders (Figure*.zip) that contain the associated data files (*.csv, *.dat) and python code notebooks (*.ipynb) for figures 3-7 in the paper. Jupyter notebook (*.ipynb) files that produce the figure files using the associated data files will run within a python environment configured with Jupyter Lab or Notebook packages.

54 ENVIRONMENTAL SCIENCES↗

NASA NTRS1976

Imaging radar observations of frozen Arctic lakes

A synthetic aperture imaging L-band radar flown aboard the NASA CV-990 remotely sensed a number of ice-covered lakes about 48 km northwest of Bethel, Alaska. The image obtained is a high resolution, two-dimensional representation of the surface backscatter cross section, and large differences in backscatter returns are observed: homogeneous low returns, homogeneous high returns and/or low returns near lake borders, and high returns from central areas. It is suggested that a low return indicates that the lake is frozen completely to the bottom, while a high return indicates the presence of fresh water between the ice cover and the lake bed.

Elachi, C.↗

NASA NTRS1987

Flow visualization studies of transverse fuel injection patterns in a nonreacting Mach 2 combustor

Planar visualization images are recorded of transverse jet mixing in a supersonic combustor flowfield, without chemical reaction, using laser-induced fluorescence from iodine molecules. Digital image processing and three-dimensional display enable complete representations of fuel penetration boundary and shock surfaces corresponding to several injection geometries and pressures.

Mcdaniel, J. C.↗

NASA NTRS2005

Inertial Motion-Tracking Technology for Virtual 3-D

In the 1990s, NASA pioneered virtual reality research. The concept was present long before, but, prior to this, the technology did not exist to make a viable virtual reality system. Scientists had theories and ideas they knew that the concept had potential, but the computers of the 1970s and 1980s were not fast enough, sensors were heavy and cumbersome, and people had difficulty blending fluidly with the machines. Scientists at Ames Research Center built upon the research of previous decades and put the necessary technology behind them, making the theories of virtual reality a reality. Virtual reality systems depend on complex motion-tracking sensors to convey information between the user and the computer to give the user the feeling that he is operating in the real world. These motion-tracking sensors measure and report an object s position and orientation as it changes. A simple example of motion tracking would be the cursor on a computer screen moving in correspondence to the shifting of the mouse. Tracking in 3-D, necessary to create virtual reality, however, is much more complex. To be successful, the perspective of the virtual image seen on the computer must be an accurate representation of what is seen in the real world. As the user s head or camera moves, turns, or tilts, the computer-generated environment must change accordingly with no noticeable lag, jitter, or distortion. Historically, the lack of smooth and rapid tracking of the user s motion has thwarted the widespread use of immersive 3-D computer graphics. NASA uses virtual reality technology for a variety of purposes, mostly training of astronauts. The actual missions are costly and dangerous, so any opportunity the crews have to practice their maneuvering in accurate situations before the mission is valuable and instructive. For that purpose, NASA has funded a great deal of virtual reality research, and benefited from the results.

Source record↗

NASA NTRS1999

An Application of Rotation- and Translation-Invariant Overcomplete Wavelets to the Registration of Remotely Sensed Imagery

A wavelet-based image registration approach has previously been proposed by the authors. In this work, wavelet coefficient maxima obtained from an orthogonal wavelet decomposition using Daubechies filters were utilized to register images in a multi-resolution fashion. Tested on several remote sensing datasets, this method gave very encouraging results. Despite the lack of translation-invariance of these filters, we showed that when using cross-correlation as a feature matching technique, features of size larger than twice the size of the filters are correctly registered by using the low-frequency subbands of the Daubechies wavelet decomposition. Nevertheless, high-frequency subbands are still sensitive to translation effects. In this work, we are considering a rotation- and translation-invariant representation developed by E. Simoncelli and integrate it in our image registration scheme. The two types of filters, Daubechies and Simoncelli filters, are then being compared from a registration point of view, utilizing synthetic data as well as data from the Landsat/ Thematic Mapper (TM) and from the NOAA Advanced Very High Resolution Radiometer (AVHRR).

LeMoigne, Jacqueline↗

NASA NTRS1999

An Application of Rotation- and Translation-Invariant Overcomplete Wavelets to the registration of Remotely Sensed Imagery

LeMoigne, Jacqueline↗

NASA NTRS1991

Representation of vegetation by continental data sets derived from NOAA-AVHRR data

Images of the normalized difference vegetation index (NDVI) are examined with specific attention given to the effect of spatial scales on the understanding of surface phenomena. A scale variance analysis is conducted on NDVI annual and seasonal images of Africa taken from 1987 NOAA-AVHRR data at spatial scales ranging from 8-512 km. The scales at which spatial variation takes place are determined and the relative magnitude of the variations are considered. Substantial differences are demonstrated, notably an increase in spatial variation with coarsening spatial resolution. Different responses in scale variance as a function of spatial resolution are noted in an analysis of maximum value composites for February and September; the difference is most marked in areas with very seasonal vegetation. The spatial variation at different scales is attributed to different factors, and methods involving the averaging of areas of transition and surface heterogeneity can oversimplify surface conditions. The spatial characteristics and the temporal variability of areas should be considered to accurately apply satellite data to global models.

Justice, C. O.↗

NASA NTRS1983

Quantile Data Analysis of Image Data

Quantile data analysis and functional statistical inference methods are introduced and applied to provide representations of spectral data which may lead to simple statistical discriminators effective for the estimation of ground truth from satellite spectral measurements. To estimate the ground truth of a pixel, the probability of each possible ground truth is estimated, given observed (estimated) quantile theoretic statistical characteristics of the multispectral image data corresponding to the pixel and its neighboring pixels. A strategy for determining which statistical characteristics discriminate best is described. Results are reported of quantile data analysis of an extensive collection of training files of image data.

Parzen, E.↗

NASA NTRS1981

Stereographic observations from geosynchronous satellites - An important new tool for the atmospheric sciences

Observations of cloud geometry using scan-synchronized stereo geostationary satellites having images with horizontal spatial resolution of approximately 0.5 km, and temporal resolution of up to 3 min are presented. The stereo does not require a cloud with known emissivity to be in equilibrium with an atmosphere with a known vertical temperature profile. It is shown that absolute accuracies of about 0.5 km are possible. Qualitative and quantitative representations of atmospheric dynamics were shown by remapping, display, and stereo image analysis on an interactive computer/imaging system. Applications of stereo observations include: (1) cloud top height contours of severe thunderstorms and hurricanes, (2) cloud top and base height estimates for cloud-wind height assignment, (3) cloud growth measurements for severe thunderstorm over-shooting towers, (4) atmospheric temperature from stereo heights and infrared cloud top temperatures, and (5) cloud emissivity estimation. Recommendations are given for future improvements in stereo observations, including a third GOES satellite, operational scan synchronization of all GOES satellites and better resolution sensors.

Hasler, A. F.↗

NASA NTRSDate not supplied

Kalman Filtering and RTS Smoothing for Arc-Jet Sample Edge Tracking

Accurate and temporally consistent measurements of test-article recession are required to characterize material response during arc-jet ablation experiments. However, image-based boundary measurements are often affected by segmentation noise, brightness variations, and frame-to-frame variability. This work extends arcjetCV [1] with filtering methods for tracking the evolving boundaries of hemispherical and ISO-Q test articles. Two boundary-tracking approaches based on Kalman filtering [2] were implemented. The first applies a point-wise Kalman filter followed by a Rauch–Tung–Striebel smoother [3] to individual boundary-point locations. The second applies the same filtering and smoothing framework to a reduced set of geometry-dependent shape parameters. The point-wise method reduces local frame-to-frame fluctuations while preserving spatial details along the detected boundary. The shape-parameter method provides a compact and geometrically constrained estimate of the sample contour. Figure 1 illustrates the two approaches for a hemispherical test article. Both methods improve temporal consistency and support more robust estimation of surface recession from arc-jet imagery. The two filtering approaches provide complementary representations of boundary evolution and improve the reliability of image-based recession measurements during arc-jet experiments.

recession measurement↗

DOE OSTI2022

The Use of Thermal Cameras for Pedestrian Detection

Visible-range camera sensors have been widely used for pedestrian detection. However, most of the methods, which employ visible-range color cameras, do not perform well under low-light and no-light conditions, e.g. during night time. Since the working principle of thermal camera sensors is mainly based on temperature and not light, they have been employed for person detection to overcome the drawbacks of visible-range sensors under these conditions. Every object gives off thermal energy, which is captured by a thermal camera sensor. When an object becomes hotter, it emits more thermal energy, and is therefore captured as much brighter or vice versa. Yet, compared to visible-range cameras, there are many additional challenges that need to be addressed when detecting pedestrians from thermal camera images. These challenges include bright hot objects close to humans, similar pixel values in an image due to weather conditions, or objects that block thermal cameras such as concrete or glass. Glass acts like a mirror for infrared radiation and reflects whatever is in front of the camera. Thus, novel methods are still required to accomplish pedestrian detection task from thermal camera images. To contribute to these efforts, we propose a new method and a modified object detection network incorporating saliency maps of thermal camera images. The features obtained from thermal images and their corresponding saliency maps are combined to obtain richer representations of pedestrian regions, and better detection performance. We perform extensive evaluations on five different datasets to compare the performance of the proposed approach with two baselines. Moreover, we evaluate and compare the transferability of these approaches by doing leave-one-out cross validation across different datasets. Furthermore, the results show that the proposed approach outperforms the baselines, and has better transferability properties across different thermal image datasets.

47 OTHER INSTRUMENTATION↗

DOE OSTI2020

Figure Descriptive Text Extraction Using Ontological Representation

Abstract Experimental research publications provide figure form resources including graphs, charts, and any type of images to effectively support and convey methods and results. To describe figures, authors add captions, which are often incomplete, and more descriptions reside in body text. This work presents a method to extract figure descriptive text from the body of scientific articles. We adopted ontological semantics to aid concept recognition of figure-related information, which generates human- and machine-readable knowledge representations from sentences. Our results show that conceptual models bring an improvement in figure descriptive sentence classification over word-based approaches.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

Physics-guided dual implicit neural representations for source separation

Significant challenges exist in efficient data analysis of most advanced experimental and observational techniques because the collected signals often include unwanted contributions, such as background and signal distortions, that can obscure the physically relevant information of interest. To address this, we have developed a self-supervised machine-learning approach for source separation using a dual implicit neural representation framework that jointly trains two neural networks: one for approximating distortions of the physical signal of interest and the other for learning the effective background contribution. Our method learns directly from the raw data by minimizing a reconstruction-based loss function without requiring labeled data or pre-defined dictionaries. We demonstrate the effectiveness of our framework by considering a challenging case study involving large-scale simulated, as well as experimental, momentum-energy-dependent inelastic neutron scattering data in a four-dimensional parameter space, characterized by heterogeneous background contributions and unknown distortions to the target signal. The method is found to successfully separate physically meaningful signals from a complex or structured background even when the signal characteristics vary across all four dimensions of the parameter space. An analytical approach that informs the choice of the regularization parameter is presented. Our method offers a versatile framework for addressing source separation problems across diverse domains, ranging from superimposed signals in astronomical measurements to structural features in biomedical image reconstructions.

47 OTHER INSTRUMENTATION↗

NASA NTRS2022

Microstructure Segmentation with Deep Learning Encoders Pre-Trained on a Large Microscopy Dataset

This study examined the improvement of microscopy segmentation accuracy by transfer learning from a large dataset of microscopy images called MicroNet. Many neural network encoder architectures, including VGG, Inception, and ResNet, were trained on over 100,000 labelled microscopy images from 54 classes. These pre-trained encoders were then embedded into multiple segmentation architectures including U-Net and DeepLabV3+ to evaluate segmentation performance on newly created benchmark microscopy datasets. Compared to ImageNet pre-training, models pre-trained on MicroNet generalized better to out-of-distribution micrographs taken under different imaging and sample conditions and were more accurate with less training data. When training with only a single Ni-superalloy image, pre-training on MicroNet produced a 72.2 percent reduction in relative segmentation error. These results suggest that transfer learning from large in-domain datasets generate models with learned feature representations that are more useful for downstream tasks and will likely improve any microscopy image analysis technique that can leverage pre-trained encoders.

machine learning↗

NASA NTRS2022

Microstructure Segmentation With Deep Learning Encoders Pre-Trained on a Large Microscopy Dataset

This study examined the improvement of microscopy segmentation intersection over union accuracy by transfer learning from a large dataset of microscopy images called MicroNet. Many neural network encoder architectures were trained on over 100,000 labeled microscopy images from 54 material classes. These pre-trained encoders were then embedded into multiple segmentation architectures including UNet and DeepLabV3+ to evaluate segmentation performance on created benchmark microscopy datasets. Compared to ImageNet pre-training, models pre-trained on MicroNet generalized better to out-of-distribution micrographs taken under different imaging and sample conditions and were more accurate with less training data. When training with only a single Ni-superalloy image, pre-training on MicroNet produced a 72.2% reduction in relative intersection over union error. These results suggest that transfer learning from large in-domain datasets generate models with learned feature representations that are more useful for downstream tasks and will likely improve any microscopy image analysis technique that can leverage pre-trained encoders.

machine learning↗

DOE OSTI2021

Adaptive machine learning for time-varying systems: low dimensional latent space tuning

Machine learning (ML) tools such as encoder-decoder convolutional neural networks (CNN) can represent incredibly complex nonlinear functions which map between combinations of images and scalars. For example, CNNs can be used to map combinations of accelerator parameters and images which are 2D projections of the 6D phase space distributions of charged particle beams as they are transported between various particle accelerator locations. Despite their strengths, applying ML to time-varying systems, or systems with shifting distributions, is an open problem, especially for large systems for which collecting new data for re-training is impractical or interrupts operations. Particle accelerators are one example of large time-varying systems for which collecting detailed training data requires lengthy dedicated beam measurements which may no longer be available during regular operations. We present a novel method of adaptive ML for time-varying systems. Our approach is to map very high (N ≈ 100k) dimensional inputs (a combination of scalar parameters and images) into the low dimensional (N ≈ 2) latent space at the output of the encoder section of an encoder-decoder CNN. We then actively tune the low dimensional latent space-based representation of complex system dynamics by the addition of an adaptively tuned feedback vector directly before the decoder sections builds back up to our image-based high-dimensional phase space density representations. This method allows us to learn correlations within and to quickly tune the characteristics of incredibly large parameter space systems and to track their evolution in real time based on feedback without massive new data sets for re-training. We demonstrate that our method can accurately predict and track the phase space of charged particle beams at various locations in a particle accelerator by adaptively adjusting in real-time while the unknown input beam distribution of the accelerator is changing in shape, charge, and offset and while the RF system of the accelerator itself is also changing in an unpredictable way. For FACET-II we demonstrate that such an approach has the potential to use transverse deflecting cavity and energy spread spectrum beam measurements to accurately predict 2D projections of the 6D phase space of the electron beam at the plasma wakefield acceleration interaction point where such diagnostics are unavailable.

47 OTHER INSTRUMENTATION↗

DOE OSTI2022

Interactively Assessing Disentanglement in GANs

Abstract Generative adversarial networks (GAN) have witnessed tremendous growth in recent years, demonstrating wide applicability in many domains. However, GANs remain notoriously difficult for people to interpret, particularly for modern GANs capable of generating photo‐realistic imagery. In this work we contribute a visual analytics approach for GAN interpretability, where we focus on the analysis and visualization of GAN disentanglement. Disentanglement is concerned with the ability to control content produced by a GAN along a small number of distinct, yet semantic, factors of variation. The goal of our approach is to shed insight on GAN disentanglement, above and beyond coarse summaries, instead permitting a deeper analysis of the data distribution modeled by a GAN. Our visualization allows one to assess a single factor of variation in terms of groupings and trends in the data distribution, where our analysis seeks to relate the learned representation space of GANs with attribute‐based semantic scoring of images produced by GANs. Through use‐cases, we show that our visualization is effective in assessing disentanglement, allowing one to quickly recognize a factor of variation and its overall quality. In addition, we show how our approach can highlight potential dataset biases learned by GANs.

Jeong, Sangwon↗

DOE OSTI2024

X-ray nano-holotomography reconstruction with simultaneous probe retrieval

In conventional tomographic reconstruction, the pre-processing step includes flat-field correction, where each sample projection on the detector is divided by a reference image taken without the sample. When using coherent X-rays as a probe, this approach overlooks the phase component of the illumination field (probe), leading to artifacts in phase-retrieved projection images, which are then propagated to the reconstructed 3D sample representation. The problem intensifies in nano-holotomography with focusing optics, which, due to various imperfections creates high-frequency components in the probe function. Here, we present a new iterative reconstruction scheme for holotomography, simultaneously retrieving the complex-valued probe function. Implemented on GPUs, this algorithm results in 3D reconstruction resolving twice thinner layers in a 3D ALD standard sample measured using nano-holotomography.

Nikitin, Viktor↗