Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “image recognition”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Image Analysis: Comparison Metrics

Despite an increased reliance on computational modeling in engineering and physics, graphically comparing computational and experimental results has often been qualitative, via a so-called ‘viewgraph norm’. Image recognition, computer vision and artificial intelligence (AI) are fields of study in themselves and rapidly progressing, but relying on AI to grade image similarity evokes a notion of asking for an expert judgement, which could be seen as an artificial version of the viewgraph norm. It is, therefore, desirable to use simpler metrics which are more tractable and unambiguous, even though they may not be as ‘intelligent.’ In this document, a few metrics are studied by systematically altering an image and their implications are analyzed

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2011

Tissue Photolithography

Tissue lithography will enable physicians and researchers to obtain macromolecules with high purity (greater than 90 percent) from desired cells in conventionally processed, clinical tissues by simply annotating the desired cells on a computer screen. After identifying the desired cells, a suitable lithography mask will be generated to protect the contents of the desired cells while allowing destruction of all undesired cells by irradiation with ultraviolet light. The DNA from the protected cells can be used in a number of downstream applications including DNA sequencing. The purity (i.e., macromolecules isolated form specific cell types) of such specimens will greatly enhance the value and information of downstream applications. In this method, the specific cells are isolated on a microscope slide using photolithography, which will be faster, more specific, and less expensive than current methods. It relies on the fact that many biological molecules such as DNA are photosensitive and can be destroyed by ultraviolet irradiation. Therefore, it is possible to protect the contents of desired cells, yet destroy undesired cells. This approach leverages the technologies of the microelectronics industry, which can make features smaller than 1 micrometer with photolithography. A variety of ways has been created to achieve identification of the desired cell, and also to designate the other cells for destruction. This can be accomplished through chrome masks, direct laser writing, and also active masking using dynamic arrays. Image recognition is envisioned as one method for identifying cell nuclei and cell membranes. The pathologist can identify the cells of interest using a microscopic computerized image of the slide, and appropriate custom software. In one of the approaches described in this work, the software converts the selection into a digital mask that can be fed into a direct laser writer, e.g. the Heidelberg DWL66. Such a machine uses a metalized glass plate (with chrome metallization) on which there is a thin layer of photoresist. The laser transfers the digital mask onto the photoresist by direct writing, with typical best resolution of 2 micrometers. The plate is then developed to remove the exposed photoresist, which leaves the exposed areas susceptible to chemical chrome etch. The etch removes the unprotected chrome. The rest of the photoresist is then removed, by either ultraviolet organic solvent or over-development. The remaining chrome pattern is quickly oxidized by atmospheric exposure (typically within 30 seconds). The ready chrome mask is now applied to the tissue slide and aligned manually, or using automatic software and pre-designed alignment marks. The slide plate sandwich is then exposed to UV to destroy the DNA of the unwanted cells. The slide and plate are separated and the slide is processed in a standard way to prepare for polymerase chain reaction (PCR) and potential identification of cancer sequences.

Wade, Lawrence A.↗
NASA NTRS2012

Ring Image Analyzer

Ring Image Analyzer software analyzes images to recognize elliptical patterns. It determines the ellipse parameters (axes ratio, centroid coordinate, tilt angle). The program attempts to recognize elliptical fringes (e.g., Newton Rings) on a photograph and determine their centroid position, the short-to-long-axis ratio, and the angle of rotation of the long axis relative to the horizontal direction on the photograph. These capabilities are important in interferometric imaging and control of surfaces. In particular, this program has been developed and applied for determining the rim shape of precision-machined optical whispering gallery mode resonators. The program relies on a unique image recognition algorithm aimed at recognizing elliptical shapes, but can be easily adapted to other geometric shapes. It is robust against non-elliptical details of the image and against noise. Interferometric analysis of precision-machined surfaces remains an important technological instrument in hardware development and quality analysis. This software automates and increases the accuracy of this technique. The software has been developed for the needs of an R&TD-funded project and has become an important asset for the future research proposal to NASA as well as other agencies.

Strekalov, Dmitry V.↗
NASA NTRS1974

Use of ERTS data for a multidisciplinary analysis of Michigan resources

The author has identified the following significant results. The results of this investigation of ratioing simulated ERTS spectral bands and several non-ERTS bands (all collected by an airborne multispectral scanner) indicate that significant terrain information is available from band-ratio images. Ratio images, which are based on the relative spectral changes which occur from one band to another, are useful for enhancing differences and aiding the image interpreter in identifying and mapping the distribution of such terrain elements as seedling crops, all bare soil, organic soil, mineral soil, forest and woodlots, and marsh areas. In addition, the ratio technique may be useful for computer processing to obtain recognition images of large areas at lower costs than with statistical decision rules. The results of this study of ratio processing of aircraft MSS data will be useful for future processing and evaluation of ERTS-1 data for soil and landform studies. Additionally, the results of ratioing spectral bands other than those currently collected by ERTS-1 suggests that some other bands (particularly a thermal band) would be useful in future satellites.

Andersen, A. L.↗
DOE OSTI2023

Differentiating multi-MeV, multi-ion spectra with CR-39 solid-state nuclear track detectors

The development of high intensity petawatt lasers has created new possibilities for ion acceleration and nuclear fusion using solid targets. In such laser-matter interaction, multiple ion species are accelerated with broad spectra up to hundreds of MeV. To measure ion yields and for species identification, CR-39 solid-state nuclear track detectors are frequently used. However, these detectors are limited in their applicability for multi-ion spectra differentiation as standard image recognition algorithms can lead to a misinterpretation of data, there is no unique relation between track diameter and particle energy, and there are overlapping pit diameter relationships for multiple particle species. In this report, we address these issues by first developing an algorithm to overcome user bias during image processing. Second, we use calibration of the detector response for protons, carbon and helium ions (alpha particles) from 0.1 to above 10 MeV and measurements of statistical energy loss fluctuations in a forward-fitting procedure utilizing multiple, differently filtered CR-39, altogether enabling high-sensitivity, multi-species particle spectroscopy. To validate this capability, we show that inferred CR-39 spectra match Thomson parabola ion spectrometer data from the same experiment. Filtered CR-39 spectrometers were used to detect, within a background of ~ 2 × 10 11 sr -1 J -1 protons and carbons, (1.3 ± 0.7) × 10 8 sr -1 J -1 alpha particles from laser-driven proton-boron fusion reactions.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
NASA NTRS2020

Assessment of holographic microscopy for quantifying marine particle size and concentration

Holographic microscopy has emerged as a tool for in situ imaging of microscopic organisms and other particles in the marine environment: appealing because of the relatively larger sampling volume and simpler optical configuration compared to other imaging systems. However, its quantitative capabilities have so far remained uncertain, in part because hologram reconstruction and image recognition have required manual operation. Here, we assess the quantitative skill of our automated hologram processing pipeline (CCV Pipeline), to evaluate the size and concentration measurements of environmental and cultured assemblages of marine plankton particles, and microspheres. Over 1 million particles, ranging from 10 to 200 μm in equivalent spherical diameter, imaged by the 4‐Deep HoloSea digital inline holographic microscope (DIHM) are analyzed. These measurements were collected in parallel with a FlowCam (FC), Imaging FlowCytobot (IFCB), and manual microscope identification. Once corrections for particle location and nonuniform illumination were developed and applied, the DIHM showed an underestimate in ESD of about 3% to 10%, but successfully reproduced the size spectral slope from environmental samples, and the size distribution of cultures (Dunaliella tertiolecta, Heterosigma akashiwo, and Prorocentrum micans) and microspheres. DIHM concentrations (order 1 to 1000 particles ml(exp −1)) showed a linear agreement (r(exp 2) = 0.73) with the other instruments, but individual comparisons at times had large uncertainty. Overall, we found the DIHM and the CCV Pipeline required extensive manual correction, but once corrected, provided concentration and size estimates comparable to the other imaging systems assessed in this study. Holographic cameras are mechanically simple, autonomous, can operate at very high pressures, and provide a larger sampling volume than comparable lens‐based tools. Thus, we anticipate that these characterization efforts will be rewarded with novel discovery in new oceanic environments.

Noah L. Walcutt↗
NASA NTRS2020

Assessment of Holographic Microscopy for Quantifying Marine Particle Size and Concentration

Holographic microscopy has emerged as a tool for in situ imaging of microscopic organisms and other particles in the marine environment: appealing because of relatively larger sampling volumes and simpler optical configurations compared to similar imaging systems. However, its quantitative capabilities have remained uncertain, in part because hologram reconstruction and image recognition have required manual operation. Here we assess the quantitative skill of our automated hologram processing pipeline, the CCV Pipeline, to evaluate the size and concentration measurements of environmental and cultured assemblages of marine plankton particles, and microspheres. Over one-million particles, ranging from 10-200 microns equivalent spherical diameter (ESD), imaged by the 4-Deep HoloSea digital inline holographic microscope (DIHM) are analyzed. These measurements were collected in parallel with FlowCam, Imaging FlowCytobot, and manual microscope identification. Once corrections for particle location and nonuniform illumination were developed and applied, the DIHM showed an underestimate in ESD of about 3-10%, but successfully reproduced the size-spectral-slope from environmental samples, and the size distribution of monocultures and microspheres. DIHM concentrations (order 1-1000 particles~ml-1) showed linear agreement (r^2=0.73) with the other instruments, but individual comparisons at times had large uncertainty.

holograpy↗
DOE OSTI2020

Image-driven discriminative and generative machine learning algorithms for establishing microstructure–processing relationships

We investigate methods of microstructure representation for the purpose of predicting processing condition from microstructure image data. A binary alloy that is currently under development as a nuclear fuel was studied for the purpose of developing an improved machine learning approach to image recognition, characterization, and building predictive capabilities linking microstructure to processing conditions. Here, we test different microstructure representations and evaluate model performance based on classification accuracy. A classification accuracy of 95.8% was achieved fordistinguishing between micrographs corresponding to ten different thermo-mechanical material processing conditions.We find that our newly developed microstructure representation describes image data well, and the traditional approachof utilizing area fractions of different phases is insufficient for distinguishing between multiple classes using a relativelysmall, imbalanced original data set of 272 images. To explore the applicability of generative methods for supplementing such limited data sets, generative adversarial networks were trained to generate artificial microstructure images. Two different generative networks were trained and tested to assess performance. Challenges and best practices associated with applying machine learning to limited microstructure image data sets is also discussed. Our work has implications for quantitative microstructure analysis, and development of microstructure-processing relationships in limited data sets typical of metallurgical process design studies.

36 MATERIALS SCIENCE↗
NASA NTRS2004

Automated Recognition of Geologically Significant Shapes in MER PANCAM and MI Images

Autonomous recognition of scientifically important information provides the capability of: 1) Prioritizing data return; 2) Intelligent data compression; 3) Reactive behavior onboard robotic vehicles. Such capabilities are desirable as mission scenarios include longer durations with decreasing interaction from mission control. To address such issues, we have implemented several computer algorithms, intended to autonomously recognize morphological shapes of scientific interest within a software architecture envisioned for future rover missions. Mars Exploration Rovers (MER) instrument payloads include a Panoramic Camera (PANCAM) and Microscopic Imager (MI). These provide a unique opportunity to evaluate our algorithms when applied to data obtained from the surface of Mars. Early in the mission we applied our algorithms to images available at the mission web site (http://marsrovers.jpl.nasa.gov/gallery/images.html), even though these are not at full resolution. Some algorithms would normally use ancillary information, e.g. camera pointing and position of the sun, but these data were not readily available. The initial results of applying our algorithms to the PANCAM and MI images are encouraging. The horizon is recognized in all images containing it; such information could be used to eliminate unwanted areas from the image prior to data transmission to Earth. Additionally, several rocks were identified that represent targets for the mini-thermal emission spectrometer. Our algorithms also recognize the layers, identified by mission scientists. Such information could be used to prioritize data return or in a decision-making process regarding future rover activities. The spherules seen in MI images were also autonomously recognized. Our results indicate that reliable recognition of scientifically relevant morphologies in images is feasible.

Morris, Robert↗
DOE OSTI2021

Computationally Efficient Multiscale Neural Networks Applied to Fluid Flow in Complex 3D Porous Media

Abstract The permeability of complex porous materials is of interest to many engineering disciplines. This quantity can be obtained via direct flow simulation, which provides the most accurate results, but is very computationally expensive. In particular, the simulation convergence time scales poorly as the simulation domains become less porous or more heterogeneous. Semi-analytical models that rely on averaged structural properties (i.e., porosity and tortuosity) have been proposed, but these features only partly summarize the domain, resulting in limited applicability. On the other hand, data-driven machine learning approaches have shown great promise for building more general models by virtue of accounting for the spatial arrangement of the domains’ solid boundaries. However, prior approaches building on the convolutional neural network (ConvNet) literature concerning 2D image recognition problems do not scale well to the large 3D domains required to obtain a representative elementary volume (REV). As such, most prior work focused on homogeneous samples, where a small REV entails that the global nature of fluid flow could be mostly neglected, and accordingly, the memory bottleneck of addressing 3D domains with ConvNets was side-stepped. Therefore, important geometries such as fractures and vuggy domains could not be modeled properly. In this work, we address this limitation with a general multiscale deep learning model that is able to learn from porous media simulation data. By using a coupled set of neural networks that view the domain on different scales, we enable the evaluation of large ( $$>512^3$$ > 512 3 ) images in approximately one second on a single graphics processing unit. This model architecture opens up the possibility of modeling domain sizes that would not be feasible using traditional direct simulation tools on a desktop computer. We validate our method with a laminar fluid flow case using vuggy samples and fractures. As a result of viewing the entire domain at once, our model is able to perform accurate prediction on domains exhibiting a large degree of heterogeneity. We expect the methodology to be applicable to many other transport problems where complex geometries play a central role.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Shape‐Dependent Multi‐Weight Magnetic Artificial Synapses for Neuromorphic Computing

Abstract In neuromorphic computing, artificial synapses provide a multi‐weight (MW) conductance state that is set based on inputs from neurons, analogous to the brain. Herein, artificial synapses based on magnetic materials that use a magnetic tunnel junction (MTJ) and a magnetic domain wall (DW) are explored. By fabricating lithographic notches in a DW track underneath a single MTJ, 3–5 stable resistance states that can be repeatably controlled electrically using spin‐orbit torque are achieved. The effect of geometry on the synapse behavior is explored, showing that a trapezoidal device has asymmetric weight updates with high controllability, while a rectangular device has higher stochasticity, but with stable resistance levels. The device data is input into neuromorphic computing simulators to show the usefulness of application‐specific synaptic functions. Implementing an artificial neural network (NN) applied to streamed Fashion‐MNIST data, the trapezoidal magnetic synapse can be used as a metaplastic function for efficient online learning. Implementing a convolutional NN for CIFAR‐100 image recognition, the rectangular magnetic synapse achieves near‐ideal inference accuracy, due to the stability of its resistance levels. This work shows MW magnetic synapses are a feasible technology for neuromorphic computing and provides design guidelines for emerging artificial synapse technologies.

Leonard, Thomas↗
DOE OSTI2021

Multiscale Modeling Meets Machine Learning: What Can We Learn?

Machine learning is increasingly recognized as a promising technology in the biological, biomedical, and behavioral sciences. There can be no argument that this technique is incredibly successful in image recognition with immediate applications in diagnostics including electrophysiology, radiology, or pathology, where we have access to massive amounts of annotated data. However, machine learning often performs poorly in prognosis, especially when dealing with sparse data. This is a field where classical physics-based simulation seems to remain irreplaceable. In this review, we identify areas in the biomedical sciences where machine learning and multiscale modeling can mutually benefit from one another: Machine learning can integrate physics-based knowledge in the form of governing equations, boundary conditions, or constraints to manage ill-posted problems and robustly handle sparse and noisy data; multiscale modeling can integrate machine learn- ing to create surrogate models, identify system dynamics and parameters, analyze sensitivities, and quantify uncertainty to bridge the scales and understand the emergence of function. With a view towards applications in the life sciences, we discuss the state of the art of combining machine learning and multiscale modeling, identify applications and opportunities, raise open questions, and address potential challenges and limitations. We anticipate that it will stimulate discussion within the community of computational mechanics and reach out to other disciplines including mathematics, statistics, computer science, artificial intelligence, biomedicine, systems biology, and precision medicine to join forces towards creating robust and efficient models for biological systems.

machine learning, multiscale modeling, physics-bas↗
DOE OSTI2021

A Review of Machine Learning Classification Using Quantum Annealing for Real-World Applications

Optimizing the training of a machine learning pipeline helps in reducing training costs and improving model performance. One such optimizing strategy is quantum annealing, which is an emerging computing paradigm that has shown potential in optimizing the training of a machine learning model. The implementation of a physical quantum annealer has been realized by D-wave systems and is available to the research community for experiments. Recent experimental results on a variety of machine learning applications using quantum annealing have shown interesting results where the performance of classical machine learning techniques is limited by limited training data and high dimensional features. This article explores the application of D-wave’s quantum annealer for optimizing machine learning pipelines for real-world classification problems. Finally, we review the application domains on which a physical quantum annealer has been used to train machine learning classifiers. We discuss and analyze the experiments performed on the D-Wave quantum annealer for applications such as image recognition, remote sensing imagery, computational biology, and particle physics. We discuss the possible advantages and the problems for which quantum annealing is likely to be advantageous over classical computation.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Collective dynamics and long-range order in thermal neuristor networks

Abstract In the pursuit of scalable and energy-efficient neuromorphic devices, recent research has unveiled a novel category of spiking oscillators, termed “thermal neuristors.” These devices function via thermal interactions among neighboring vanadium dioxide resistive memories, emulating biological neuronal behavior. Here, we show that the collective dynamical behavior of networks of these neurons showcases a rich phase structure, tunable by adjusting the thermal coupling and input voltage. Notably, we identify phases exhibiting long-range order that, however, does not arise from criticality, but rather from the time non-local response of the system. In addition, we show that these thermal neuristor arrays achieve high accuracy in image recognition and time series prediction through reservoir computing, without leveraging long-range order. Our findings highlight a crucial aspect of neuromorphic computing with possible implications on the functioning of the brain: criticality may not be necessary for the efficient performance of neuromorphic systems in certain computational tasks.

Science & Technology - Other Topics↗
DOE OSTI2023

Quantum machine learning with differential privacy

Abstract Quantum machine learning (QML) can complement the growing trend of using learned models for a myriad of classification tasks, from image recognition to natural speech processing. There exists the potential for a quantum advantage due to the intractability of quantum operations on a classical computer. Many datasets used in machine learning are crowd sourced or contain some private information, but to the best of our knowledge, no current QML models are equipped with privacy-preserving features. This raises concerns as it is paramount that models do not expose sensitive information. Thus, privacy-preserving algorithms need to be implemented with QML. One solution is to make the machine learning algorithm differentially private, meaning the effect of a single data point on the training dataset is minimized. Differentially private machine learning models have been investigated, but differential privacy has not been thoroughly studied in the context of QML. In this study, we develop a hybrid quantum-classical model that is trained to preserve privacy using differentially private optimization algorithm. This marks the first proof-of-principle demonstration of privacy-preserving QML. The experiments demonstrate that differentially private QML can protect user-sensitive information without signficiantly diminishing model accuracy. Although the quantum model is simulated and tested on a classical computer, it demonstrates potential to be efficiently implemented on near-term quantum devices [noisy intermediate-scale quantum (NISQ)]. The approach’s success is illustrated via the classification of spatially classed two-dimensional datasets and a binary MNIST classification. This implementation of privacy-preserving QML will ensure confidentiality and accurate learning on NISQ technology.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

On-policy learning-based deep reinforcement learning assessment for building control efficiency and stability

Artificial intelligence technologies have emerged as a game changer not only in specific applications such as image recognition and machine translation but also in many scientific domains. In particular, as deep reinforcement learning (DRL) has shown great success in complex control problems, DRL-based control has been considered as a potential solution to efficiently control and manage building systems. However, broad assessment of DRL-based building control is still required to characterize their pros and cons in comparison with conventional building control methods (e.g., rule-based feedback controls). In this paper, we assessed DRL-based controls with on-policy learning-based algorithms and continuous control actions for cooling control of large office buildings in the summer season to minimize whole-building energy use and occupant discomfort. We compared DRL-based control methods with two baseline control methods: (1) a pre-determined schedule with supply temperature and static pressure setpoints, and (2) advanced reset method that adjusts setpoints based on heuristic rules, i.e., ASHRAE Guideline 36. We also tested the DRL algorithms to evaluate their performances in multiple climate locations. We found that DRL-based control methods outperformed the baseline control methods in terms of energy savings while maintaining a thermal comfort. DRL reduced energy use between ~4%–22% on average compared to the baseline methods, depending on climate location. We also evaluated DRL-based control in terms of control stability and showed that DRL-based methods should address the span of hardware lifetimes in practical operations.

control stability↗
DOE OSTI2023

Differentiable Earth mover’s distance for data compression at the high-luminosity LHC

Abstract The Earth mover’s distance (EMD) is a useful metric for image recognition and classification, but its usual implementations are not differentiable or too slow to be used as a loss function for training other algorithms via gradient descent. In this paper, we train a convolutional neural network (CNN) to learn a differentiable, fast approximation of the EMD and demonstrate that it can be used as a substitute for computing-intensive EMD implementations. We apply this differentiable approximation in the training of an autoencoder-inspired neural network (encoder NN) for data compression at the high-luminosity LHC at CERN The goal of this encoder NN is to compress the data while preserving the information related to the distribution of energy deposits in particle detectors. We demonstrate that the performance of our encoder NN trained using the differentiable EMD CNN surpasses that of training with loss functions based on mean squared error.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗