Search NASA⌕ Search

SEARCH · Search NASA

Results for “3d imaging”

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 415 records · Page 23
All sourcesNASA NTRSDOE OSTI
NASA NTRSDate not supplied

3D Localization of Defects in Facility Inspection

Wind tunnels are crucial facilities that support the aerospace industry. However, these facilities are large, complex, and pose unique maintenance and inspection requirements. Manual inspections to identify defects such as cracks, missing fasteners, leaks, and foreign objects are important but labor and schedule intensive. Our goal is to utilize small Unmanned Aircraft Systems (sUAS) and computer vision-based analysis to automate the inspection of the interior and exterior of NASA’s critical wind tunnel facilities. We detect missing fasteners as our defect class, and detect existing fasteners to provide potential future missing fastener sites for preventative maintenance. These detections are done on both 2D raw images and in 3D space to provide a visual reference and real world location to facilitate repairs. A dataset was created consisting of images taken along a grid-like pattern of an interior tunnel section in the AEDC National Full-Scale Aerodynamics Complex (NFAC) at NASA Ames Research Center. Our method uses object detection to create image level bounding boxes of the fasteners and missing fasteners, then uses photogrammetry to create a mapping from 2D image locations to 3D real world locations. The image level bounding boxes and the 2D to 3D mapping are then combined to determine the 3D location of the defects. We describe the data collection, photogrammetry, and computer vision techniques used for object detection as well as a quantitative analysis of the method.

Small Unmanned Aircraft Systems (sUAS)↗
DOE OSTI2020

Feature‐shared adaptive‐boost deep learning for invasiveness classification of pulmonary subsolid nodules in CT images

Purpose In clinical practice, invasiveness is an important reference indicator for differentiating the malignant degree of subsolid pulmonary nodules. These nodules can be classified as atypical adenomatous hyperplasia (AAH), adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), or invasive adenocarcinoma (IAC). The automatic determination of a nodule's invasiveness based on chest CT scans can guide treatment planning. However, it is challenging, owing to the insufficiency of training data and their interclass similarity and intraclass variation. To address these challenges, we propose a two‐stage deep learning strategy for this task: prior‐feature learning followed by adaptive‐boost deep learning. Methods The adaptive‐boost deep learning is proposed to train a strong classifier for invasiveness classification of subsolid nodules in chest CT images, using multiple 3D convolutional neural network (CNN)‐based weak classifiers. Because ensembles of multiple deep 3D CNN models have a huge number of parameters and require large computing resources along with more training and testing time, the prior‐feature learning is proposed to reduce the computations by sharing the CNN layers between all weak classifiers. Using this strategy, all weak classifiers can be integrated into a single network. Results Tenfold cross validation of binary classification was conducted on a total of 1357 nodules, including 765 noninvasive (AAH and AIS) and 592 invasive nodules (MIA and IAC). Ablation experimental results indicated that the proposed binary classifier achieved an accuracy of with an AUC of 81.3 . These results are superior compared to those achieved by three experienced chest imaging specialists who achieved an accuracy of , , and , respectively. About 200 additional nodules were also collected. These nodules covered 50 cases for each category (AAH, AIS, MIA, and IAC, respectively). Both binary and multiple classifications were performed on these data and the results demonstrated that the proposed method definitely achieves better performance than the performance achieved by nonensemble deep learning methods. Conclusions It can be concluded that the proposed adaptive‐boost deep learning can significantly improve the performance of invasiveness classification of pulmonary subsolid nodules in CT images, while the prior‐feature learning can significantly reduce the total size of deep models. The promising results on clinical data show that the trained models can be used as an effective lung cancer screening tool in hospitals. Moreover, the proposed strategy can be easily extended to other similar classification tasks in 3D medical images.

Wang, Jun↗
NASA NTRS2017

Imaging Systems for Size Measurements of Debrisat Fragments

The overall objective of the DebriSat project is to provide data to update existing standard spacecraft breakup models. One of the key sets of parameters used in these models is the physical dimensions of the fragments (i.e., length, average-cross sectional area, and volume). For the DebriSat project, only fragments with at least one dimension greater than 2 mm are collected and processed. Additionally, a significant portion of the fragments recovered from the impact test are needle-like and/or flat plate-like fragments where their heights are almost negligible in comparison to their other dimensions. As a result, two fragment size categories were defined: 2D objects and 3D objects. While measurement systems are commercially available, factors such as measurement rates, system adaptability, size characterization limitations and equipment costs presented significant challenges to the project and a decision was made to develop our own size characterization systems. The size characterization systems consist of two automated image systems, one referred to as the 3D imaging system and the other as the 2D imaging system. Which imaging system to use depends on the classification of the fragment being measured. Both imaging systems utilize point-and-shoot cameras for object image acquisition and create representative point clouds of the fragments. The 3D imaging system utilizes a space-carving algorithm to generate a 3D point cloud, while the 2D imaging system utilizes an edge detection algorithm to generate a 2D point cloud. From the point clouds, the three largest orthogonal dimensions are determined using a convex hull algorithm. For 3D objects, in addition to the three largest orthogonal dimensions, the volume is computed via an alpha-shape algorithm applied to the point clouds. The average cross-sectional area is also computed for 3D objects. Both imaging systems have automated size measurements (image acquisition and image processing) driven by the need to quickly and accurately measure tens of thousands of debris fragments. Moreover, the automated size measurement reduces potential fragment damage/mishandling and ability for accuracy and repeatability. As the fragment characterization progressed, it became evident that the imaging systems had to be revised. For example, an additional view was added to the 2D imaging system to capture the height of the 2D object. This paper presents the DebriSat project's imaging systems and calculation techniques in detail; from design and development to maturation. The experiences and challenges are also shared.

Shiotani, B.↗
NASA NTRS2009

3D Lunar Terrain Reconstruction from Apollo Images

Generating accurate three dimensional planetary models is becoming increasingly important as NASA plans manned missions to return to the Moon in the next decade. This paper describes a 3D surface reconstruction system called the Ames Stereo Pipeline that is designed to produce such models automatically by processing orbital stereo imagery. We discuss two important core aspects of this system: (1) refinement of satellite station positions and pose estimates through least squares bundle adjustment; and (2) a stochastic plane fitting algorithm that generalizes the Lucas-Kanade method for optimal matching between stereo pair images.. These techniques allow us to automatically produce seamless, highly accurate digital elevation models from multiple stereo image pairs while significantly reducing the influence of image noise. Our technique is demonstrated on a set of 71 high resolution scanned images from the Apollo 15 mission

Broxton, Michael J.↗
NASA NTRS1988

A knowledge-based object recognition system for applications in the space station

A knowledge-based three-dimensional (3D) object recognition system is being developed. The system uses primitive-based hierarchical relational and structural matching for the recognition of 3D objects in the two-dimensional (2D) image for interpretation of the 3D scene. At present, the pre-processing, low-level preliminary segmentation, rule-based segmentation, and the feature extraction are completed. The data structure of the primitive viewing knowledge-base (PVKB) is also completed. Algorithms and programs based on attribute-trees matching for decomposing the segmented data into valid primitives were developed. The frame-based structural and relational descriptions of some objects were created and stored in a knowledge-base. This knowledge-base of the frame-based descriptions were developed on the MICROVAX-AI microcomputer in LISP environment. The simulated 3D scene of simple non-overlapping objects as well as real camera data of images of 3D objects of low-complexity have been successfully interpreted.

Dhawan, Atam P.↗
DOE OSTI2021

End-to-end orientation estimation from 2D cryo-EM1images

Cryo-electron microscopy (cryo-EM) is a Nobel Prize-winning technique for deter-mining high-resolution 3D structures of biological macromolecules. A 3D structure is reconstructed from hundreds of thousands of noisy 2D projection images. However, existing 3D reconstruction methods are still time-consuming, and one of the major computational bottlenecks is to recover the unknown orientation of the particle in16each 2D image. The dominant methods typically exploit expensive global search on each image to estimate the missing orientations. Here, a novel end-to-end supervised learning method is introduced to directly recover the missing orientations from 2D cryo-EM images. A neural network is used to approximate the mapping from images to orientations. Furthermore, a robust loss function is proposed for optimizing the parameters of the network, which can handle both asymmetric and symmetric 3D structures. Experiments on synthetic datasets with various symmetry types confirm that the neural network is capable of recovering orientations from 2D cryo-EM images, and the results on one real cryo-EM dataset further demonstrate its potential in more challenging imaging conditions.

3D reconstruction↗
NASA NTRS2006

Synthesis of 3D Model of a Magnetic Field-Influenced Body from a Single Image

A method for recovery of a 3D model of a cloud-like structure that is in motion and deforming but approximately governed by magnetic field properties is described. The method allows recovery of the model from a single intensity image in which the structure's silhouette can be observed. The method exploits envelope theory and a magnetic field model. Given one intensity image and the segmented silhouette in the image, the method proceeds without human intervention to produce the 3D model. In addition to allowing 3D model synthesis, the method's capability to yield a very compact description offers further utility. Application of the method to several real-world images is demonstrated.

Wang, Cuilan↗
NASA NTRS2024

Direct Imaging of Magnetohydrodynamic Wave Mode Conversion Near A 3D Null Point on the Sun

Mutual conversion of various kinds of magnetohydrodynamic (MHD) waves can have profound impacts on wave propagation, energy transfer, and heating of the solar chromosphere and corona. Mode conversion occurs when an MHD wave travels through a region where the Alfvén and sound speeds are equal (e.g., a 3D magnetic null point). Here we report the direct extreme ultraviolet (EUV) imaging of mode conversion from a fast-mode to a slow-mode MHD wave near a 3D null point using Solar Dynamics Observatory/Atmospheric Imaging Assembly (SDO/AIA) observations. An incident fast EUV wavefront associated with an adjacent eruptive flare propagates laterally through a neighboring pseudostreamer. Shortly after the passage of the fast EUV wave through the null point, a slow-mode wave appears near the null that propagates upward along the open structures and simultaneously downward along the separatrix encompassing the fan loops of the pseudostreamer base. These observations suggest the existence of mode conversion near 3D nulls in the solar corona, as predicted by theory and MHD simulations. Moreover, we observe decaying transverse oscillations in both the open and closed structures of the pseudostreamer, along with quasiperiodic type III radio bursts indicative of repetitive episodes of electron acceleration.

Pankaj Kumar↗
DOE OSTI2020

How Low Can You Go? Using Synthetic 3D Imagery to Drastically Reduce Real-World Training Data for Object Detection

Deep convolutional neural networks (DCNNs) currently provide state-of-the-art performance on image classification and object detection tasks, and there are many global security mission areas where such models could be extremely useful. Crucially, the success of these models is driven in large part by the widespread availability of high-quality open source data sets such as Image Net, Common Objects in Context (COCO), and KITTI, which contain millions of images with thousands of unique labels. However, global security relevant objects-of-interest can be difficult to obtain: relevant events are low frequency and high consequence; the content of relevant images is sensitive; and adversaries and proliferators seek to obscure their activities. For these cases where exemplar data is hard to come-by, even fine-tuning an existing model with available data can be effectively impossible. Recent work demonstrated that models can be trained using a combination of real-world and synthetic images generated from 3D representations; that such models can exceed the performance of models trained using real-world data alone; and that the generated images need not be perfectly realistic (Tremblay, et al., 2018). However, this approach still required hundreds to thousands of real-world images for training and fine tuning, which for sparse, global security-relevant datasets can be an unrealistic hurdle. In this research, we validate the performance and behavior of DCNN models as we drive the number of real-world images used for training object detection tasks down to a minimal set. We perform multiple experiments to identify the best approach to train DCNNs from an extremely small set of real-world images. In doing so, we: Develop state-of-the-art, parameterized 3D models based on real-world images and sample from their parameters to increase the variance in synthetic image training data; Use machine learning explainability techniques to highlight and correct through targeted training the biases that result from training using completely synthetic images; and Validate our results by comparing the performance of the models trained on synthetic data to one another, and to a control model created by fine-tuning an existing ImageNet-trained model with a limited number (hundreds) of real-world images.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2013

Building a 2.5D Digital Elevation Model from 2D Imagery

When projecting imagery into a georeferenced coordinate frame, one needs to have some model of the geographical region that is being projected to. This model can sometimes be a simple geometrical curve, such as an ellipse or even a plane. However, to obtain accurate projections, one needs to have a more sophisticated model that encodes the undulations in the terrain including things like mountains, valleys, and even manmade structures. The product that is often used for this purpose is a Digital Elevation Model (DEM). The technology presented here generates a high-quality DEM from a collection of 2D images taken from multiple viewpoints, plus pose data for each of the images and a camera model for the sensor. The technology assumes that the images are all of the same region of the environment. The pose data for each image is used as an initial estimate of the geometric relationship between the images, but the pose data is often noisy and not of sufficient quality to build a high-quality DEM. Therefore, the source imagery is passed through a feature-tracking algorithm and multi-plane-homography algorithm, which refine the geometric transforms between images. The images and their refined poses are then passed to a stereo algorithm, which generates dense 3D data for each image in the sequence. The 3D data from each image is then placed into a consistent coordinate frame and passed to a routine that divides the coordinate frame into a number of cells. The 3D points that fall into each cell are collected, and basic statistics are applied to determine the elevation of that cell. The result of this step is a DEM that is in an arbitrary coordinate frame. This DEM is then filtered and smoothed in order to remove small artifacts. The final step in the algorithm is to take the initial DEM and rotate and translate it to be in the world coordinate frame [such as UTM (Universal Transverse Mercator), MGRS (Military Grid Reference System), or geodetic] such that it can be saved in a standard DEM format and used for projection.

Padgett, Curtis W.↗
DOE OSTI2020

Fractal breakage of porous carbonate sand particles: Microstructures and mechanisms

In situ, three-dimensional (3D) characterizations of particle breakage in porous carbonate sands are presented, for the first time, with synchrotron-based micro computed tomography. Evolution of grain-scale characteristics are identified and quantified via elaborate image processing and topology analyses. The sequential 3D images reveal distinctly different fracture mechanisms for carbonate sands from silica sands. The angular shape of carbonate sand particles facilitates bending fracture, and particles with a lower sphericity and a higher porosity are more prone to break. 3D crack networks extracted from fractured particles imply considerable cleavage along initial pores. The fractal dimension of crack networks increases with external loading due to crack branching via cleavage. The resultant fragment size distribution also appears fractal and the fractal feature is valid down to the breakage limit of calcium carbonate. Crack propagation along the initial pores reduces the energy barrier for particle breakage and thus fracture strength of particles

Carbonate sands↗
DOE OSTI2020

Development of a predictive capability of short-pulse laser-driven broadband x-ray radiography

High intensity, short-pulse laser interaction with a solid metal target produces broadband hard x-rays potentially for various applications of x-ray radiography. In this work, experimental benchmarking of numerical modelling for short-pulse laser-driven broadband x-ray radiography is presented. Angular dependent x-ray spectra are first calculated with a hybrid particle-in-cell code, Large Scale Plasma (LSP), using fast electron parameters inferred from an analysis of measured bremsstrahlung signals. Subsequently, a calculated x-ray spectrum in the direction of radiography is used in photon transport calculations using a Monte Carlo code, Particle and Heavy Ion Transport code System (PHITS), to simulate a radiographic image including a modelled 3D test object, an x-ray attenuation filter and an image plate detector. Simulated radiographic images are compared with measurements obtained in an experiment using a 50-TW Leopard short-pulse laser at the University of Nevada Reno. Results show that simulations reproduce the experimental images well for three different attenuation filters (plastic, aluminium, and brass), while 1D transmission profiles for the plastic and aluminium filters are quantitatively in good agreement. The modelling approach established in this work could be used as a predictive tool to simulate radiographic images of complex 3D solid objects at any arbitrary angular position or to optimize experimental components such as the source spectrum, x-ray attenuation filters and a detector type depending on a radiographic object without carrying out radiographic experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

NEURAL NETWORK FOR COHERENT DIFFRACTION IMAGE INVERSION

A deep neural network model plus automatic differentiation is developed for retrieving phase information from 3D coherent diffraction images. The model is implemented using Tensorflow and the training dataset is generated using physics-based atomistic simulations. Custom codes are written to handle the resampling of diffraction images to oversampling ratios appropriate for the neural network model.

CHAN, HENRY↗
DOE OSTI2022

3D Time-Lapse Electrical Resistivity Imaging of Rock Damage Patterns and Gas Flow Paths Resulting from Two Underground Chemical Explosions

Abstract Rock damage from underground nuclear explosions (UNEs) has a strong influence on sub-surface gas movement and on seismic waveform characteristics, both of which are used to detect UNEs. Although advanced numerical simulation capabilities exist to predict rock damage patterns and corresponding detection signals, those predictions are dependent on (generally) unknown properties of the host rock. For example, the effects of in-situ mechanical heterogeneities on the explosively generated damage/fractures that provide gas flow pathways to the surface are not well understood, due largely to the difficulty in accessing and characterizing the near-source region. In this paper we demonstrate the emerging use of electrical resistivity tomography (ERT) for imaging rock damage and gas flow patterns resulting from two relatively small-scale underground chemical explosions. Pre-explosion ERT and crosshole seismic imaging revealed a natural fracture zone within the test bed. Post-explosion imaging revealed that the damage zone was non-symmetric and was focused primarily within the pre-existing fracture zone, located 10 m above the first explosion and 5 m above the second explosion. Time-lapse ERT imaging of heated air injected into the detonation borehole revealed the primary gas flow paths to be within the upper margin of the same primary damage zone. These results point to the utility of ERT imaging for understanding rock damage and gas flow patterns under experimental conditions, and to the importance of understanding the effects of geologic heterogeneity on UNE detection signals, particularly gas surface breakthrough times.

58 GEOSCIENCES↗
DOE OSTI2020

Method for measurement of TRISO kernel and layer volumes by X-ray computed tomography

Layer dimensions are key parameters for as-fabricated tristructural-isotropic (TRISO) particle fuel as well as for post-irradiation examination of particle performance. Layer thicknesses are typically measured by optical microscopy of the particle cross section near mid-plane, while layer volumes are estimated with serial sectioning and microscopy. This method for measuring layer volumes is limited due to the resolution limit imposed by slice thickness and the effect of the polishing process on delicate irradiated particle microstructures. In this study, image processing software has been developed to segregate the three-dimensional (3D) TRISO particle images provided by x-ray computed tomography (XCT) into high-resolution volumetric data for the kernel, all particle layers, and any internal voids introduced during irradiation. These data can be used to analyze key in-reactor behaviors of TRISO particles such as kernel swelling or buffer shrinkage, which are important inputs for fuel performance modeling.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Scintillation event imaging with a single photon avalanche diode camera

Position and time measurements of scintillation events encode information about the radiation source. Single photon avalanche diode (SPAD) arrays offer multiple-megapixel spatial resolution and tens of picoseconds temporal resolution for detecting single photons. Current lensless designs for measuring scintillation events use sensors that are lower in spatial resolution. Camera-based designs use sensors that are lower in temporal resolution or readout rate and cannot image individual interactions. Here we propose to image scintillation events in a thick, monolithic scintillator using a high-resolution SPAD camera. We demonstrate that a commercial SPAD camera is able to gather sufficient signal to image individual scintillation events and observe 3D shifts in their spatial distribution. Simulations show that a SPAD camera can localize individual scintillation events in 3D. We report direct imaging of gamma-ray interactions in a scintillator with a SPAD camera. The proposed design may allow to measure complex signatures of individual particles interacting in the scintillator.

47 OTHER INSTRUMENTATION↗
NASA NTRS2012

Time-Domain Terahertz Computed Axial Tomography NDE System

NASA has identified the need for advanced non-destructive evaluation (NDE) methods to characterize aging and durability in aircraft materials to improve the safety of the nation's airline fleet. 3D THz tomography can play a major role in detection and characterization of flaws and degradation in aircraft materials, including Kevlar-based composites and Kevlar and Zylon fabric covers for soft-shell fan containment where aging and durability issues are critical. A prototype computed tomography (CT) time-domain (TD) THz imaging system has been used to generate 3D images of several test objects including a TUFI tile (a thermal protection system tile used on the Space Shuttle and possibly the Orion or similar capsules). This TUFI tile had simulated impact damage that was located and the depth of damage determined. The CT motion control gan try was designed and constructed, and then integrated with a T-Ray 4000 control unit and motion controller to create a complete CT TD-THz imaging system prototype. A data collection software script was developed that takes multiple z-axis slices in sequence and saves the data for batch processing. The data collection software was integrated with the ability to batch process the slice data with the CT TD-THz image reconstruction software. The time required to take a single CT slice was decreased from six minutes to approximately one minute by replacing the 320 ps, 100-Hz waveform acquisition system with an 80 ps, 1,000-Hz waveform acquisition system. The TD-THZ computed tomography system was built from pre-existing commercial off-the-shelf subsystems. A CT motion control gantry was constructed from COTS components that can handle larger samples. The motion control gantry allows inspection of sample sizes of up to approximately one cubic foot (.0.03 cubic meters). The system reduced to practice a CT-TDTHz system incorporating a COTS 80- ps/l-kHz waveform scanner. The incorporation of this scanner in the system allows acquisition of 3D slice data with better signal-to-noise using a COTS scanner rather than the gchirped h scanner. The system also reduced to practice a prototype for commercial CT systems for insulating materials where safety concerns cannot accommodate x-ray. A software script was written to automate the COTS software to collect and process TD-THz CT data.

Zimdars, David↗