SEARCH · Search NASA

Results for “Image processing algorithm”

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 1,765 records · Page 98

All sources NASA NTRS DOE OSTI

DOE OSTI2026

Data Summarization and Inference at Scale

This is the final report for the DOE ASCR grant SC-0022260, Data Summarization and Inference at Scale, PI: Alex Pothen, Purdue University. The goal of the project was to solve data-intensive and compute-intensive problems in the physical sciences, engineering, information science, data science, etc. by designing and implementing new algorithms that could work with a subset of the data. The four subgoals were: (a) The solution of problems where the data is too large to be stored in the memory of a computer. In this streaming model of computation, the data arrives as a stream of elements to the computer, each element is processed as it arrives, and a decision is made to discard the data or to store it; only a small subset of the data proportional to the size of the output solution is stored, and when all the data has been streamed, a solution to the problem is computed from the stored subset. (b) The use of machine learning methods to compute solutions to data-intensive problems. The use of GPUs is critical to obtain high performance on machine learning tasks, but their memory sizes are smaller relative to that of CPUs. For large-scale problems, the data is sampled many times, and small samples are used with repetition, for robustness, to compute solutions to inference tasks. This sampling reduces the memory required to solve the problem, but attention is needed to avoid slow convergence to the solutions, and reduced accuracy of inference. We propose submodular optimization, Large Language Models, and physics-informed neural networks to enable GPU computations here. (c) Modeling and visualization of high-dimensional data using interpretable features. Clinical proteomic data sets from immunology for the detection of cancer and other diseases are temporal and high-dimensional, and algorithms for visualizing these data sets using clinically interpretable features are lacking. We propose methods that compute distances based on the optimal transportation problem and graph edit distances to address this problem. We also propose the use of optimal transport-based distances, spatial statistics, and network structure to classify image data sets, We apply these algorithms to electron micrographs of the peripheral nervous system in the digestive tract. (d) The design of data-intensive algorithms on emerging architectures, specifically, noisy, intermediate-scale quantum (NISQ) devices. Quantum computers offer the possibility of exploring large solution spaces due to the principle of superposition, but current quantum computers are limited by few qubits, short coherence times due to noise, poor interconections among the qubits, etc. We propose the use of the divide and conquer paradigm to solve large-scale problems, wherein collections of small subproblems are solved on the quantum devices, and the solutions to the subproblems are integrated into a solution for the original problem on a classical computer.

97 MATHEMATICS AND COMPUTING↗

NASA NTRSDate not supplied

The Dark Target aerosol retrieval algorithm applied to Low Earth Orbit and GEOstationary imagers: progress towards an integrated LEO-GEO view of global aerosol

The relatively simple dark-target (DT) aerosol retrieval algorithm provides products of spectral aerosol optical depth (AOD) from measurements of multi-spectral reflectance in visible, near-infrared and shortwave infrared wavelength bands. Originally developed for Moderate-resolution Imaging Spectroradiometer (MODIS aboard Terra and Aqua) in Low-Earth Orbit (LEO), DT has been ported to Visible Infrared Imaging Suite (VIIRS aboard Suomi-NPP and NOAA-20, also in LEO), to enhanced-MODIS Airborne Simulator (eMAS, on an airborne platform), and now to sensors in GEOstationary orbit (Advanced Himawari Imager - AHI aboard Himawari-8 and Advanced Baseline Imagers – ABI aboard GOES-16 and 17). Together, these new datasets not only extend upon the 20+ year MODIS aerosol record, but also expand the temporal sampling and/or spatial resolution. Between July and October of 2019, NASA participated in two field experiments on opposite sides of the globe. These included FIREX-AQ which focused on fire and smoke in the Western U.S., and then CAMP2EX which targeted aerosol/cloud interactions around the Philippines. We have performed DT aerosol retrievals on all images from all sensors during these three months, validated against ground observations from stationary and mobile sunphotometer sites, and have begun to develop a synergy that represents semi-global observations every half hour. The resulting aerosol products are being used as context and for model assimilation, thus providing the framework for more complete characterization of global aerosol transport and lifecycle. Here, we report on progress, as well as remaining challenges such as data management, computer processing, and accounting for differences between GEO and LEO observation geometry and surface reflectance parameterization.

dark target↗

DOE OSTI2023

Manifold projection image segmentation for nano-XANES imaging

As spectral imaging techniques are becoming more prominent in science, advanced image segmentation algorithms are required to identify appropriate domains in these images. We present a version of image segmentation called manifold projection image segmentation (MPIS) that is generally applicable to a broad range of systems without the need for training because MPIS uses unsupervised machine learning with a few physically motivated hyperparameters. We apply MPIS to nanoscale x-ray absorption near edge structure (XANES) imaging, where XANES spectra are collected with nanometer spatial resolution. We show the superiority of manifold projection over linear transformations, such as the commonly used principal component analysis (PCA). Moreover, MPIS maintains accuracy while reducing computation time and sensitivity to noise compared to the standard nano-XANES imaging analysis procedure. Finally, we demonstrate how multimodal information, such as x-ray fluorescence data and spatial location of pixels, can be incorporated into the MPIS framework. We propose that MPIS is adaptable for any spectral imaging technique, including scanning transmission x-ray microscopy, where the length scale of domains is larger than the resolution of the experiment.

36 MATERIALS SCIENCE↗

NASA NTRS2009

NASA Tech Briefs, October 2009

Topics covered include: Light-Driven Polymeric Bimorph Actuators; Guaranteeing Failsafe Operation of Extended-Scene Shack-Hartmann Wavefront Sensor Algorithm; Cloud Water Content Sensor for Sounding Balloons and Small UAVs; Pixelized Device Control Actuators for Large Adaptive Optics; T-Slide Linear Actuators; G4FET Implementations of Some Logic Circuits; Electrically Variable or Programmable Nonvolatile Capacitors; System for Automated Calibration of Vector Modulators; Complementary Paired G4FETs as Voltage-Controlled NDR Device; Three MMIC Amplifiers for the 120-to-200 GHz Frequency Band; Low-Noise MMIC Amplifiers for 120 to 180 GHz; Using Ozone To Clean and Passivate Oxygen-Handling Hardware; Metal Standards for Waveguide Characterization of Materials; Two-Piece Screens for Decontaminating Granular Material; Mercuric Iodide Anticoincidence Shield for Gamma-Ray Spectrometer; Improved Method of Design for Folding Inflatable Shells; Ultra-Large Solar Sail; Cooperative Three-Robot System for Traversing Steep Slopes; Assemblies of Conformal Tanks; Microfluidic Pumps Containing Teflon[Trademark] AF Diaphragms; Transparent Conveyor of Dielectric Liquids or Particles; Multi-Cone Model for Estimating GPS Ionospheric Delays; High-Sensitivity GaN Microchemical Sensors; On the Divergence of the Velocity Vector in Real-Gas Flow; Progress Toward a Compact, Highly Stable Ion Clock; Instruments for Imaging from Far to Near; Reflectors Made from Membranes Stretched Between Beams; Integrated Risk and Knowledge Management Program -- IRKM-P; LDPC Codes with Minimum Distance Proportional to Block Size; Constructing LDPC Codes from Loop-Free Encoding Modules; MMICs with Radial Probe Transitions to Waveguides; Tests of Low-Noise MMIC Amplifier Module at 290 to 340 GHz; and Extending Newtonian Dynamics to Include Stochastic Processes.

Source record↗

NASA NTRS2003

Three-dimensional ocular kinematics during eccentric rotations: evidence for functional rather than mechanical constraints

Previous studies have reported that the translational vestibuloocular reflex (TVOR) follows a three-dimensional (3D) kinematic behavior that is more similar to visually guided eye movements, like pursuit, rather than the rotational VOR (RVOR). Accordingly, TVOR rotation axes tilted with eye position toward an eye-fixed reference frame rather than staying relatively fixed in the head like in the RVOR. This difference arises because, contrary to the RVOR where peripheral image stability is functionally important, the TVOR like pursuit and saccades cares to stabilize images on the fovea. During most natural head and body movements, both VORs are simultaneously activated. In the present study, we have investigated in rhesus monkeys the 3D kinematics of the combined VOR during yaw rotation about eccentric axes. The experiments were motivated by and quantitatively compared with the predictions of two distinct hypotheses. According to the first (fixed-rule) hypothesis, an eye-position-dependent torsion is computed downstream of a site for RVOR/TVOR convergence, and the combined VOR axis would tilt through an angle that is proportional to gaze angle and independent of the relative RVOR/TVOR contributions to the total eye movement. This hypothesis would be consistent with the recently postulated mechanical constraints imposed by extraocular muscle pulleys. According to the second (image-stabilization) hypothesis, an eye-position-dependent torsion is computed separately for the RVOR and the TVOR components, implying a processing that takes place upstream of a site for RVOR/TVOR convergence. The latter hypothesis is based on the functional requirement that the 3D kinematics of the combined VOR should be governed by the need to keep images stable on the fovea with slip on the peripheral retina being dependent on the different functional goals of the two VORs. In contrast to the fixed-rule hypothesis, the data demonstrated a variable eye-position-dependent torsion for the combined VOR that was different for synergistic versus antagonistic RVOR/TVOR interactions. Furthermore, not only were the eye-velocity tilt slopes of the combined VOR as much as 10 times larger than what would be expected based on extraocular muscle pulley location, but also eye velocity during antagonistic RVOR/TVOR combinations often tilted opposite to gaze. These results are qualitatively and quantitatively consistent with the image-stabilization hypothesis, suggesting that the eye-position-dependent torsion is computed separately for the RVOR and the TVOR and that the 3D kinematics of the combined VOR are dependent on functional rather than mechanical constraints.

Non-NASA Center↗

DOE OSTI2023

Automated detection of double nuclei galaxies using GOTHIC and the discovery of a large sample of dual AGN

We present a novel algorithm to detect double nuclei galaxies (DNG) called Gothic (Graph-bOosTed iterated HIll Climbing) – that detects whether a given image of a galaxy has two or more closely separated nuclei. Our aim is to test for the presence of dual/multiple active galactic nuclei (AGN) in galaxies that visually represent a DNG. Although galaxy mergers are common, the detection of dual AGN is rare. Their detection is very important as they help us understand the formation of supermassive black hole (SMBH) binaries, SMBH growth and AGN feedback effects in multiple nuclei systems. There is thus a need for an algorithm to do a systematic survey of existing imaging data for the discovery of DNGs and dual AGNs. We have tested Gothic on an established sample of DNGs with a 100 per cent detection rate and subsequently conducted a blind search of 1 million SDSS DR16 galaxies (with spectroscopic data available) lying in the redshift range of z = 0 to 0.75. From the list of candidate DNGs found, we have detected 159 dual AGNs, of which 2 are triple AGN systems. Our results show that dual AGNs are not common, and triple AGN even rarer. The colour (u–r) magnitude plots of the DNGs indicate that star formation is quenched as the nuclei come closer and as the AGN fraction increases. The quenching is especially prominent for dual/triple AGN galaxies that lie at the extreme end of the red sequence.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2023

Foundations of machine learning for low-temperature plasmas: methods and case studies

Abstract Machine learning (ML) and artificial intelligence have proven to be an invaluable tool in tackling a vast array of scientific, engineering, and societal problems. The main drivers behind the recent proliferation of ML in practically all aspects of science and technology can be attributed to: (a) improved data acquisition and inexpensive data storage; (b) exponential growth in computing power; and (c) availability of open-source software and resources that have made the use of state-of-the-art ML algorithms widely accessible. The impact of ML on the field of low-temperature plasmas (LTPs) could be particularly significant in the emerging applications that involve plasma treatment of complex interfaces in areas ranging from the manufacture of microelectronics and processing of quantum materials, to the LTP-driven electrification of the chemical industry, and to medicine and biotechnology. This is primarily due to the complex and poorly-understood nature of the plasma-surface interactions in these applications that pose unique challenges to the modeling, diagnostics, and predictive control of LTPs. As the use of ML is becoming more prevalent, it is increasingly paramount for the LTP community to be able to critically analyze and assess the concepts and techniques behind data-driven approaches. To this end, the goal of this paper is to provide a tutorial overview of some of the widely-used ML methods that can be useful, amongst others, for discovering and correlating patterns in the data that may be otherwise impractical to decipher by human intuition alone, for learning multivariable nonlinear data-driven prediction models that are capable of describing the complex behavior of plasma interacting with interfaces, and for guiding the design of experiments to explore the parameter space of plasma-assisted processes in a systematic and resource-efficient manner. We illustrate the utility of various supervised, unsupervised and active learning methods using LTP datasets consisting of commonly-available, information-rich measurements (e.g. optical emission spectra, current–voltage characteristics, scanning electron microscope images, infrared surface temperature measurements, Fourier transform infrared spectra). All the ML demonstrations presented in this paper are carried out using open-source software; the datasets and codes are made publicly available. The FAIR guiding principles for scientific data management and stewardship can accelerate the adoption and development of ML in the LTP community.

Physics↗

NASA NTRS1995

Investigation of biomass burning and aerosol loading and transport in South America utilizing geostationary satellites

This study attempts to assess the extent of burning and associated aerosol transport regimes in South America and the South Atlantic using geostationary satellite observations, in order to explore the possible roles of biomass burning in climate change and more directly in atmospheric chemistry and radiative transfer processes. Modeling and analysis efforts have suggested that the direct and indirect radiative effects of aerosols from biomass burning may play a major role in the radiative balance of the earth and are an important factor in climate change calculations. One of the most active regions of biomass burning is located in South America, associated with deforestation in the selva (forest), grassland management, and other agricultural practices. As part of the NASA Aerosol Interdisciplinary Program, we are utilizing GOES-7 (1988) and GOES-8 (1995) visible and multispectral infrared data (4, 11, and 12 microns) to document daily biomass burning activity in South America and to distinguish smoke/aerosols from other multi-level clouds and low-level moisture. This study catalogues the areal extent and transport of smoke/aerosols throughout the region and over the Atlantic Ocean for the 1988 (July-September) and 1995 (June-October) biomass burning seasons. The smoke/haze cover estimates are compared to the locations of fires to determine the source and verify the haze is actually associated with biomass burning activities. The temporal resolution of the GOES data (half-hourly in South America) makes it possible to determine the prevailing circulation and transport of aerosols by considering a series of visible and infrared images and tracking the motion of smoke, haze and adjacent clouds. The study area extends from 40 to 70 deg W and 0 to 40 deg S with aerosol coverage extending over the Atlantic Ocean when necessary. Fire activity is estimated with the GOES Automated Biomass Burning Algorithm (ABBA). To date, our efforts have focused on GOES-7 and GOES-8 ABBA development, algorithm development for aerosol monitoring, data acquisition and archiving, and participation in the SCAR-C and SCAR-B field programs which have provided valuable information for algorithm testing and validation. Implementation of the initial version of the GEOS-8 ABBA on case studies in North, Central, and South America has demonstrated the improved capability for monitoring diurnal fire activity and smoke/aerosol transport with the GOES-8 throughout the Western Hemisphere.

Menzel, Paul↗

NASA NTRSDate not supplied

Using Machine Learning to Infer Material Properties of Debris Fragments from X-ray Images in the DebriSat Project

The DebriSat project is a collaboration effort with the NASA Orbital Debris Program Office, the U.S. Space Force Space Systems Command Center, The Aerospace Corporation, and the University of Florida. To date, over 200,000 fragments from this ground-based, hypervelocity impact experiment have been collected, and processing is underway to determine their physical characteristics, such as material, shape, color, characteristic length, and average cross-sectional area. The x-ray process is primarily used to identify the location of the fragments and estimated size for extraction, so that these physical characteristics can be assessed. This paper proposes a machine learning-based approach to characterize materials from x-ray images of debris fragments embedded in soft-catch foam used in the DebriSat project. The novel methodology discussed in this paper will highlight the use of x-ray imagery data to characterize these fragments without extraction or a human-in-the-loop. Both supervised and unsupervised machine learning techniques are utilized with this approach to infer the physical parameters of the fragments embedded in the soft-catch foam panels used in the impact experiment based on x-ray images of the foam panels. Additionally, 3D reconstructions of the extracted fragments are created with images taken from two different angles using the structure from motion (SfM) method. The characteristic lengths and shape from the 3D reconstruction, alongside the physical characteristics of the debris, are used in the inference of the material type. To develop and test the approach, a dataset of x-ray images of debris fragments of varying sizes and materials is collected. Supervised learning methods such as convolutional neural networks (CNNs), support vector machines (SVM), decision trees, and random forest classifiers are used due to the high-dimensional feature spaces of the debris and nonlinear decision boundaries for material categorization. Given the limited pre-labeled data of embedded debris materials smaller than 10 mm, unsupervised machine learning techniques such as clustering algorithms and autoencoders are used, in addition to supervised learning methods. The clustering algorithms group similar fragments together based on their physical properties, and autoencoders reduce the dimensionality of the x ray images and extract relevant features. The performance of the proposed approach's is analyzed using a range of statistical methods, including confusion matrices, receiver operating characteristic curves, and precision-recall curves. The results are compared with those obtained using a baseline approach that relies on manual identification and classification of debris fragments. To evaluate the effectiveness of different machine learning methods, statistical tests such as t-tests, ANOVA, and cross-validation are performed, comparing the performance of CNNs, SVMs, clustering algorithms, and autoencoders. Additional analysis needs to be conducted to identify any sources of bias or variability that may affect the results, such as variations in imaging conditions or fragmentation patterns. Other topics explored are limitations, refinements, and the potential use of semi-supervised learning techniques, such as self-training to label unlabeled datasets and co-training using x-ray images taken from two different angles as two different models.

Saik Anam Siam↗

NASA NTRS2023

Using Machine Learning to Infer Material Properties of Debris Fragments from X-ray Images in the DebriSat Project

Saik Anam Siam↗

DOE OSTI2018

Machine Learning Algorithms for Matching Theories, Simulations, and Observations in Cosmology (Final Project)

The search for simple and elegant equations to explain and predict observable natural phenomena has been a driving force in scientific theory for several centuries. The great successes of Newton's laws, Maxwell's equations, the Schrodinger equation, and the theory of General Relativity have encouraged this perspective, and perhaps a beautiful equation comprising a “theory of everything” remains to be discovered. But now scientists are grappling with systems of ever-increasing complexity, fraught with nonlinearity, and with the quantities of interest often only indirectly observed. Nature may resist a simple description, and the most important discoveries of the next century may be complex theories with countless variables and parameters. The era of big data opens up a promising new approach to scientific discovery in this setting. The predictions of modern theories, even if complex or nonlinear, can be examined through detailed and computationally-intensive simulations that take days or weeks to run and fill petabytes of disk space. The upcoming generation of scientific instruments will provide petabytes of observations at unprecedented resolution and depth. The challenge is to compare data to simulations so as to test theories and identify those that best match the observations. The data are numerous and rich but are subject to noise and various systematic biases. The simulations are large and detailed but are so costly that the effective sample size is small. Moreover, the mapping from data to simulation to the parameters of interest is typically complicated and ill-conditioned. Although these challenges, and the solutions we propose, apply across the sciences, we will ground our efforts by focusing on cosmology. Here, we are truly entering the era of big data in several senses. Ever larger experiments survey larger and larger cosmological volumes, resulting in enormous amounts of pixel data to process and extract the information of cosmological interest. These include large imaging surveys, which collect all the light in broad wavelength ranges over significant fractions of the sky, and spectroscopic surveys, which measure the light as a function of wavelength to map out 3D distances. These experiments are designed to shed light on the big unanswered questions in cosmology. Distilling the petabytes of data into observed quantities for comparison with theory is a daunting task. But these large experiments do not only result in large amounts of data to analyze: if we wish to compare to theories of physics, then we need increasingly extensive predictions for the observed quantities as well, which requires ever-larger cosmological simulations that take substantial -- and at times prohibitive -- computational resources to produce. Our goal is to develop statistical and machine learning methods for using observed and simulated data to advance machine learning with applications cosmology. In particular, we focus our research efforts on the following tasks: The challenge for automated science is that it is computationally impossible and statistically dangerous to consider every possible model in order to find the best one. We will develop Bayesian Optimization based active-learning methods that accelerate both the execution of the simulations and the search for best-fitting parameters. The key idea is to make the simulations adaptive -- across resolutions, time, and parameters -- using the data to search as the simulation runs; Most ML methods operate on simple finite dimensional feature vectors. However, many cosmology and other science applications require ML methods that can operate on more complex objects as inputs or outputs such as functions, distributions, or sets and point clouds. Our goal is to develop efficient ML methods for this problem and demonstrate their applicability in Cosmology, Astrophysics, and other science problems.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2025

Multi-frequency progressive refinement for learned inverse scattering

Interpreting scattered acoustic and electromagnetic wave patterns is a computational task that enables remote imaging in a number of important applications, including medical imaging, geophysical exploration, sonar and radar detection, and nondestructive testing of materials. However, accurately and stably recovering an inhomogeneous medium from far-field scattered wave measurements is a computationally difficult problem, due to the nonlinear and non-local nature of the forward scattering process. We design a neural network, called Multi-Frequency Inverse Scattering Network (MFISNet), and a training method to approximate the inverse map from far-field scattered wave measurements at multiple frequencies. We consider three variants of MFISNet, with the strongest performing variant inspired by the recursive linearization method — a commonly used technique for stably inverting scattered wavefield data — that progressively refines the estimate with higher frequency content. MFISNet outperforms past methods in regimes with high-contrast, heterogeneous large objects, and inhomogeneous unknown backgrounds.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2023

Learning Canonical Embeddings for Unsupervised Shape Correspondence With Locally Linear Transformations

We present a new approach to unsupervised shape correspondence learning between pairs of point clouds. We make the first attempt to adapt the classical locally linear embedding algorithm (LLE)-originally designed for nonlinear dimensionality reduction-for shape correspondence. The key idea is to find dense correspondences between shapes by first obtaining high-dimensional neighborhood-preserving embeddings of low-dimensional point clouds and subsequently aligning the source and target embeddings using locally linear transformations. We demonstrate that learning the embedding using a new LLE-inspired point cloud reconstruction objective results in accurate shape correspondences. More specifically, the approach comprises an end-to-end learnable framework of extracting high-dimensional neighborhood-preserving embeddings, estimating locally linear transformations in the embedding space, and reconstructing shapes via divergence measure-based alignment of probability density functions built over reconstructed and target shapes. Our approach enforces embeddings of shapes in correspondence to lie in the same universal/canonical embedding space, which eventually helps regularize the learning process and leads to a simple nearest neighbors approach between shape embeddings for finding reliable correspondences. Comprehensive experiments show that the new method makes noticeable improvements over state-of-the-art approaches on standard shape correspondence benchmark datasets covering both human and nonhuman shapes.

deformation↗

NASA NTRS2003

Transferring Knowledge from Observations and Models to Decision Makers: An Overview and Challenges

Over the last 25 years, a tremendous progress has been made in the Earth science space-based remote sensing observations, technologies and algorithms. Such advancements have improved the predictability by providing lead-time and accuracy of forecast in weather, climate, natural hazards, and natural resources. It has further reduced or bounded the overall uncertainties by partially improving our understanding of planet Earth as an integrated system that is governed by non-linear and chaotic behavior. Many countries such US, European Community, Japan, China and others have invested billions of dollars in developing and launching space-based assets in the low earth (LEO) and geostationary (GEO) orbits. However, the wealth of this scientific knowledge that has potential of extracting monumental socio-economic benefits from such large investments have been slow in reaching to public and decision makers. For instance, there are a number of areas such as energy forecasting, aviation safety, agricultural competitiveness, disaster management, security, air quality and public health can directly take advantage. Nevertheless, we all live in a global economy that depends on access to the best available Earth Science information for all inhabitants of this planet. This paper surveys and examines a number such applications in terms of their architecture, maturity and economic applicability as they apply to the societal needs. A detailed analysis is also presented of various challenges and issues that pertain to a number of areas such as: (1) difficulties in making a speedy transition of data and information from observations and models to relevant Decision Support Systems (DSS) or tools, (2) data and models inter-operability issues, (3) limitations of spatial, spectral and temporal resolution, (4) communication limitations as dictated by the availability of image processing and data compression techniques. Additionally, the most critical element amongst all is the organizational and management boundaries that must be resolved at local, state, national and international levels to implement and realize free flow of such vital information. This paper also makes attempts to address this topic and discuss possible approaches to deal with this quandary.

Habib, Shahid↗

DOE OSTI2024

Real World Use Case Evaluation of Radar Retro-reflectors for Autonomous Vehicle Lane Detection Applications

Lane detection plays a critical role in autonomous vehicles for safe and reliable navigation. Lane detection is traditionally accomplished using a camera sensor and computer vision processing. The downside of this traditional technique is that it can be computationally intensive when high quality images at a fast frame rate are used and has reliability issues from occlusion such as, glare, shadows, active road construction, and more. This study addresses these issues by exploring alternative methods for lane detection in specific scenarios caused from road construction-induced lane shift and sun glare. Specifically, a U-Net, a convolutional network used for image segmentation, camera-based lane detection method is compared with a radar-based approach using a new type of sensor previously unused in the autonomous vehicle space: radar retro-reflectors. This evaluation is performed using ground truth data, obtained by measuring the lane positions and transforming them into pixel coordinates. The performance of each method is assessed using the statistical R2 score, indicating the correlation between the detected lane lines and the ground truth. The results show that the U-Net camera-based method exhibits limitations in accurately detecting and aligning the lane lines, particularly in challenging scenarios. However, the radar-based lane detection method demonstrates a strong correlation with the ground truth which implies that the use of this sensor may improve current reliability issues from conventional camera lane detection approach. Furthermore, the study highlights the limitations of the U-Net model for camera lane detection, especially in scenarios with sun glare. This study shows that infrastructure-based radar retro-reflectors can improve autonomous vehicle lane detection reliability. The integration of different sensor modalities and the development of advanced computer vision algorithms are crucial for improving the accuracy, reliability, and energy efficiency of lane detection systems. Addressing these challenges contributes to the advancement of autonomous vehicles and the realization of safer and more efficient transportation systems.

Brown, Nicolas E.↗

DOE OSTI2019

Generalized, Energy-conserving Numerical Simulations of Particles in General Relativity. II. Test Particles in Electromagnetic Fields and GRMHD

Observations of compact objects, in the form of radiation spectra, gravitational waves by LIGO/Virgo, and direct imaging with the Event Horizon Telescope, are currently the main information sources on plasma physics in extreme gravity. Modeling such physical phenomena Requires numerical methods that allow for the simulation of microscopic plasma dynamics in the presence of both strong gravity and electromagnetic fields. In Bacchini et al. (2018), we presented a detailed study of numerical techniques for the integration of free geodesic motion. In this paper, we extend the study by introducing electromagnetic forces in the simulation of charged particles in curved spacetimes. We extend the Hamiltonian energy-conserving method presented in Bacchini et al. to include the Lorentz force, and we test its performance compared to that of standard explicit Runge–Kutta and implicit midpoint rule schemes against analytic solutions. Then, we show the application of the numerical schemes to the integration of test particle trajectories in general relativistic magnetohydrodynamic (GRMHD) simulations by modifying the algorithms to handle grid-based electromagnetic fields. We test this approach by simulating ensembles of charged particles in a static GRMHD configuration obtained with the black hole accretion code (BHAC).

79 ASTRONOMY AND ASTROPHYSICS↗

NASA NTRS2022

A Comprehensive Machine Learning Study to Classify Precipitation Type over Land from Global Precipitation Measurement Microwave Imager (GPM-GMI) Measurements

Precipitation type is a key parameter used for better retrieval of precipitation characteristics as well as to understand the cloud–convection–precipitation coupling processes. Ice crystals and water droplets inherently exhibit different characteristics in different precipitation regimes (e.g., convection, stratiform), which reflect on satellite remote sensing measurements that help us distinguish them. The Global Precipitation Measurement (GPM) Core Observatory’s microwave imager (GMI) and dual-frequency precipitation radar (DPR) together provide ample information on global precipitation characteristics. As an active sensor, the DPR provides an accurate precipitation type assignment, while passive sensors such as the GMI are traditionally only used for empirical understanding of precipitation regimes. Using collocated precipitation type flags from the DPR as the “truth”, this paper employs machine learning (ML) models to train and test the predictability and accuracy of using passive GMI-only observations together with ancillary information from a reanalysis and GMI surface emissivity retrieval products. Out of six ML models, four simple ones (support vector machine, neural network, random forest, and gradient boosting) and the 1-D convolutional neural network (CNN) model are identified to produce 90–94% prediction accuracy globally for five types of precipitation (convective, stratiform, mixture, no precipitation, and other precipitation), which is much more robust than previous similar effort. One novelty of this work is to introduce data augmentation (subsampling and bootstrapping) to handle extremely unbalanced samples in each category. A careful evaluation of the impact matrices demonstrates that the polarization difference (PD), brightness temperature (Tc) and surface emissivity at high-frequency channels dominate the decision process, which is consistent with the physical understanding of polarized microwave radiative transfer over different surface types, as well as in snow and liquid clouds with different microphysical properties. Furthermore, the view-angle dependency artifact that the DPR’s precipitation flag bears with does not propagate into the conical-viewing GMI retrievals. This work provides a new and promising way for future physics-based ML retrieval algorithm development.

machine learning/artificial intelligence↗

NASA NTRS2002

Relationship Between Surface Reflectance in the Visible and Mid-IR used in MODIS Aerosol Algorithm-Theory

Data from the Moderate Resolution Imaging Spectroradiometer (MODIS) instrument that flies in polar orbit on the Terra platform, are used to derive the aerosol optical thickness and properties over land and ocean. The relationships between visible reflectance (at blue, rho(sub blue), and red, rho(sub red)) and mid-infrared (at 2.1 microns, rho(sub 2.1)) are used in the MODIS aerosol retrieval algorithm to derive global distribution of aerosols over the land. These relations have been established from a series of measurements indicating that rho(sub blue) is approximately 0.5 rho(sub red) is approximately 0.25 rho(sub 2.1). Here we use a model to describe the transfer of radiation through a vegetation canopy composed of randomly oriented leaves to assess the theoretical foundations for these relationships. Calculations for a wide range of leaf area indices and vegetation fractions show that rho(sub blue) is consistently about 1/4 of rho(sub 2.1) as used by MODIS for the whole range of analyzed cases, except for very dark soils, such as those found in burn scars. For its part, the ratio rho(sub red)/rho(sub 2.1) varies from less than the empirically derived value of 1/2 for dense and dark vegetation, to more than 1/2 for bright mixture of soil and vegetation. This is in agreement with measurements over uniform dense vegetation, but not with measurements over mixed dark scenes. In the later case the discrepancy is probably mitigated by shadows due to uneven canopy and terrain on a large scale. It is concluded that the value of this ratio should ideally be made dependent on the land cover type in the operational processing of MODIS data, especially over dense forests.

Kaufman, Yoram J.↗