Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “image understanding”

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 91 records · Page 5

The Geospace Dynamics Observatory; a Mission of Discovery for Geospace

A few examples of potential advances include: 1. Unparalleled advances in the connection of the upper atmosphere to the Sun. In the aurora and lower latitudes, extending the duration of uninterrupted images would advance understanding of the transfer of energy from the Sun to the upper atmosphere and the response of the space environment. 2. Advances in the influence of waves and tides on the upper atmosphere. Increasing both the signal to noise and the duration ofthe observations would reveal contributions that are not identifiable using other approaches. 3. The ability to probe the mechanisms that control the evolution of planetary atmospheres. The vantage point provided by this mission allows the flux of hydrogen (which is tied to the escape of water from a planet) to be mapped globally. It also allows unique observations of changes in the atmospheric structure and their causes.

Spann, James↗

Physical State and Distribution of Materials at the Surface of Pluto from New Horizons LEISA Imaging Spectrometer

From Earth based observations Pluto is known to be the host of N2, CH4 and CO ices and also a dark red material. Very limited spatial distribution information is available from rotational visible and near-infrared spectral curves obtained from hemispheric measurements. In July 2015 the New Horizons spacecraft reached Pluto and its satellite system and recorded a large set of data. The LEISA spectro-imager of the RALPH instruments are dedicated to the study of the composition and physical state of the materials composing the surface. In this paper we report a study of the distribution and physical state of the ices and non-ice materials on Pluto's illuminated surface and their mode and degree of mixing. Principal Component analysis as well as various specific spectral indicators and correlation plots are used on the first set of 2 high resolution spectro-images from the LEISA instrument covering the whole illuminated face of Pluto at the time of the New Horizons encounter. Qualitative distribution maps have been obtained for the 4 main condensed molecules, N2, CH4, CO, H2O as well as for the visible-dark red material. Based on specific spectral indicators, using either the strength or the position of absorption bands, these 4 molecules are found to indicate the presence of 3 different types of ices: N2-rich:CH4:CO ices, CH4-rich(:CO:N2?) ices and H2O ice. The mixing lines between these ices and with the dark red material are studied using scatter plots between the various spectral indicators. CH4 is mixed at the molecular level with N2, most probably also with CO, thus forming a ternary molecular mixture that follows its phase diagram with low solubility limits. The occurrence of a N2-rich - CH4-rich ices mixing line associated with a progressive decrease of the CO/CH4 ratio tells us that a fractionation sublimation sequence transforms one type of ice to the other forming either a N2-rich - CH4-rich binary mixture at the surface or an upper CH4-rich ice crust that may hide the N2-rich ice below. The strong CH4-rich - H2O mixing line witnesses the subsequent sublimation of the CH4-rich ice lag left behind by the N2:CO sublimation (N spring-summer), or a direct condensation of CH4 ice on the cold H2O ice (S autumn). The weak mixing line between CH4-containing ices and the dark red material and the very sharp spatial transitions between these ices and this non-volatile material are probably due to thermal incompatibility. Finally the occurrence of a H2O ice - red material mixing line advocates for a spatial mixing of the red material covering H2O ice, with possibly a small amount intimately mixed in water ice. From this analysis of the different materials distribution and their relative mixing lines, H2O ice appears to be the substratum on which other ices condense or non-volatile organic material is deposited from the atmosphere. N2-rich ices seem to evolve to CH4-dominated ices, possibly still containing traces of CO and N2, as N2 and CO sublimate away. The spatial distribution of these materials is very complex. The high spatial definition of all these composition maps, as well as those at even higher resolution that will be soon available, will allow us to compare them with Pluto's geologic features observed by LORRI panchromatic and MVIC multispectral imagers to better understand the geophysical processes in action at the surface of this astonishingly active frozen world.

Schmitt, B.↗

Transforming Our Understanding of the X-ray Universe: The Imaging X-ray Polarimeter Explorer (IXPE)

Accurate X-ray polarimetry can provide unique information on high-energy-astrophysical processes and sources. As there have been no meaningful X-ray polarization measurements of cosmic sources since our pioneering work in the 1970's, the time is ripe to explore this new parameter space in X-ray astronomy. To accomplish this requires a well-calibrated and well understood system that-particularly for an Explorer mission-has technical, cost, and schedule credibility. The system that we shall present satisfies these conditions, being based upon completely calibrated imaging- and polarization-sensitive detectors and proven X-ray-telescope technology.

Weisskopf, Martin C.↗

On the problem of embedding picture elements in regions

A few new algorithms for region growing in pictures were developed. These algorithms are a step toward finding a satisfactory solution to the image segmentation problem, and in-depth understanding of the problems of nonsemantic image segmentation. The algorithms utilize a sequential decision approach for region boundaries detection. The sequential decisions are supported by a stochastic algorithm that maintains local statistics of the region near the boundaries as the region grows. A few illustrations of the algorithm's performance are included.

Yakimovsky, Y.↗

Concept-based Analysis of Neural Networks via Vision-Language Models

The analysis of vision-based deep neural networks (DNNs) is highly desirable but it is very challenging due to the difficulty of expressing formal specifications for vision tasks and the lack of efficient verification procedures. In this paper, we propose to leverage emerging multimodal, vision-language, foundation models (VLMs) as a lens through which we can reason about vision models. VLMs have been trained on a large body of images accompanied by their textual description, and are thus implicitly aware of high-level, human-understandable concepts describing the images. We describe a logical specification language Con spec designed to facilitate writing specifications in terms of these concepts. To define and formally check Con spec specifications, we build a map between the internal representations of a given vision model and a VLM, leading to an efficient verification procedure of natural-language properties for vision models. We demonstrate our techniques on a ResNet-based classifier trained on the RIVAL-10 dataset using CLIP as the multimodal model.

Large Vision Language Models↗

Some characteristics and advantages of LANDSAT 3 return beam vidicon images

One advantage of a LANDSAT 3 return beam vidicon (RBV) scene is the synoptic view it provides. A 1:500,000 paper print of the size 19.4 cm x 19.4 cm encompasses an area of about 3,721 square miles (9,637 sq. km.) and this compares very favorably in coverage to small scale aerial photography. A second important advantage is that the ground resolution is 38 by 38 meters. Compared to the resolution of the MSS which is 79 by 79 meters the RBV provides a wealth of detail useful for many purposes. A great variety of man made structures are visible including bridges, roads, docks, and small airfields. Agricultural field patterns are clear and settlements of various sizes can be identified and delineated. A third advantage is related to the approximate panchromatic spectral response displayed in RBV images. Since many aerial photographic interpreters are accustomed to working with panchromatic photographs, they should have no difficulty understanding and putting RBV images immediately into use.

Baker, S.↗

High Resolution Three-Color Imaging of Spirals With Nuclear Star-Forming Rings

Nuclear rings in barred spirals offer an opportunity to study starburst properties in order to develop an understanding of the evolution of star formation in galaxies. To achieve this understanding, a large scale imaging survey in the H alpha line and in the B and I broad bands has been performed. Analysis of all galaxies that reveal nuclear rings in the H alpha line will be compared to numerical models so that the relative ages between the starforming clumps can be estimated. The luminosity function of the starforming regions will be related to the measured properties of the associated star-cluster and the required ionizing flux. Also B - I color index images will be performed to indicate the location of the dust lanes.

Mazzuca, Lisa↗

Computational Diagnostic Techniques for Electromagnetic Scattering: Analytical Imaging, Near Fields, and Surface Currents

This paper presents three techniques and the graphics implementations which can be used as diagnostic aides in the design and understanding of scattering structures: Imaging, near fields, and surface current displays. The imaging analysis is a new bistatic k space approach which has potential for much greater information than standard experimental approaches. The near field and current analysis are implementations of standard theory while the diagnostic graphics displays are implementations exploiting recent computer engineering work station graphics libraries.

Hom, Kam W.↗

Global Imaging Mission

Recent correlative observations have advanced our understanding of the solar-terrestrial environment. Among the suite of useful measurements, global auroral images have proven to be useful in guiding global models. Global auroral images provide an irreplaceable contextual measure that is required for understanding the momentum and mass transfer between the magnetosphere and ionosphere, and required for understanding their dynamic response to the solar wind. Global auroral images can provide much more than contextual information. Quantitative data, such as hemispheric power input distribution on a global scale provides needed information with regards to the dynamic nature of the magnetosphere. However, truly global images (both hemispheres simultaneously) with appropriate temporal, spectral and spatial resolution are needed in order to fully comprehend the dynamic nature of the magnetosphere. Differences in response of the two hemispheres as a function of time, space, and energy, provide needed boundary conditions for models as well as a gateway to understand the mechanism relevant to the provide a quantitative measure of the magnetosphere, A mission concept based on the need to provide a quantitative measure of the global distribution of the power input and average energy including in-situ particle measurement will be described.

Lumnierzheim, D.↗

Analysis and use of VAS satellite data

Four interrelated investigations have examined the analysis and use of VAS satellite data. A case study of VAS-derived mesoscale stability parameters suggested that they would have been a useful supplement to conventional data in the forecasting of thunderstorms on the day of interest. A second investigation examined the roles of first guess and VAS radiometric data in producing sounding retrievals. Broad-scale patterns of the first guess, radiances, and retrievals frequently were similar, whereas small-scale retrieval features, especially in the dew points, were often of uncertain origin. Two research tasks considered 6.7 micron middle tropospheric water vapor imagery. The first utilized radiosonde data to examine causes for two areas of warm brightness temperature. Subsidence associated with a translating jet streak was important. The second task involving water vapor imagery investigated simulated imagery created from LAMPS output and a radiative transfer algorithm. Simulated image patterns were found to compare favorably with those actually observed by VAS. Furthermore, the mass/momentum fields from LAMPS were powerful tools for understanding causes for the image configurations.

Fuelberg, Henry E.↗

Plasmaspheric Structures Observed by the CLUSTER and IMAGE Spacecraft

Global EUV imaging observations by the IMAGE satellite have revealed spectacularly complex changes in plasmaspheric structures as the plasmaspheric plasmas respond to geomagnetic activity while remaining under varying degrees of influence by co-rotation, depending on the radial distance. This complex plasmaspheric plasma dynamics, with its numerous scales of variability, is clearly far from being well understood. There is now renewed interest in studying the plasmasphere due to its apparent connections with the development of the ring current and radiation belt and loss of ionospheric plasmas. Earlier in the mission, the Cluster spacecraft only crossed the plasmapause (L - 4) occasionally and made measurements of the cold plasma in the plasmasphere and plasmaspheric drainage plumes. The study by Darrouzet et al. [2006], for example, provided detailed analyses of in situ Cluster observations and IMAGE EUV imaging observations of three plasmaspheric plumes detected in April-June, 2002. Within the next couple of years, Cluster orbit will change, causing perigee to migrate to lower altitudes, and thus providing excellent opportunities to obtain more detailed measurements of the plasmasphere. In this paper, we report our analyses of the earlier Cluster-IMAGE events by incorporating the different perspectives provided by the IMAGE Radio Plasma Imager (RPI) observations through the plasmasphere. We will describe our new understanding of the Cluster-IMAGE events and their implications on plasmaspheric dynamics.

Fung, S. F.↗

The Shuttle Imaging Radar B (SIR-B) experiment report

The primary objective of the SIR-B experiment was to acquire multiple-incidence-angle radar imagery of a variety of Earth's surfaces to better understand the effects of imaging geometry on radar backscatter. A complementary objective was to map extensive regions of particular interest. Under these broad objectives, many specific scientific experiments were defined by the 43 SIR-B Science Team members, including studies in the area of geology, vegetation, radar penetration, oceanography, image analysis, and calibration technique development. Approximately 20 percent of the planned digital data were collected, meeting 40 percent of the scientific objectives. This report is an overview of the SIR-B experiment and includes the science investigations, hardware design, mission scenario, mission operations, events of the actual missions, astronaut participation, data products (including auxiliary data), calibrations, and a summary of the actual coverage. Also included are several image samples.

Cimino, Jo Bea↗

Stage Cylindrical Immersive Display

Panoramic images with a wide field of view intend to provide a better understanding of an environment by placing objects of the environment on one seamless image. However, understanding the sizes and relative positions of the objects in a panorama is not intuitive and prone to errors because the field of view is unnatural to human perception. Scientists are often faced with the difficult task of interpreting the sizes and relative positions of objects in an environment when viewing an image of the environment on computer monitors or prints. A panorama can display an object that appears to be to the right of the viewer when it is, in fact, behind the viewer. This misinterpretation can be very costly, especially when the environment is remote and/or only accessible by unmanned vehicles. A 270 cylindrical display has been developed that surrounds the viewer with carefully calibrated panoramic imagery that correctly engages their natural kinesthetic senses and provides a more accurate awareness of the environment. The cylindrical immersive display offers a more natural window to the environment than a standard cubic CAVE (Cave Automatic Virtual Environment), and the geometry allows multiple collocated users to simultaneously view data and share important decision-making tasks. A CAVE is an immersive virtual reality environment that allows one or more users to absorb themselves in a virtual environment. A common CAVE setup is a room-sized cube where the cube sides act as projection planes. By nature, all cubic CAVEs face a problem with edge matching at edges and corners of the display. Modern immersive displays have found ways to minimize seams by creating very tight edges, and rely on the user to ignore the seam. One significant deficiency of flat-walled CAVEs is that the sense of orientation and perspective within the scene is broken across adjacent walls. On any single wall, parallel lines properly converge at their vanishing point as they should, and the sense of perspective within the scene contained on only one wall has integrity. Unfortunately, parallel lines that lie on adjacent walls do not necessarily remain parallel. This results in inaccuracies in the scene that can distract the viewer and subtract from the immersive experience of the CAVE.

Abramyan, Lucy↗

SAR imaging of ocean waves - Theory

A SAR imaging integral for a rough surface is derived. Aspects of distributed target imaging and questions of ocean-wave imaging are considered. A description is presented of the results of analyses which are performed on aircraft and a spacecraft data in order to gain an understanding of the SAR imaging of ocean waves. The analyzed data illustrate the effect of radar resolution on the images of azimuthally traveling waves, the dependence of image distortion on the angle which the waves make with the radar flight path, and the dependence of the focusing parameter of the radar matched filter on the ocean wave period for azimuthally traveling waves. A dependence of ocean-wave modulation on significant wave height is also observed. The observed dependence of the modulations of azimuth waves on radar resolution is in contradiction to the hypothesis that these modulations are caused mainly by velocity bunching.

Jain, A.↗

3d Particle Geometry Characterization of Apollo Regolith Samples Via Laser Diffraction and Dynamic Image Analysis

Establishing a fundamental understanding of regolith mechanics hinges on the characterization of the foundational physical parameters that dictate granular particle interactions both on micro- and macro-scales. Specifically, particle size distribution, particle geometry and grain density govern the geotechnical properties, including compressibility, shear strength, hydraulic conductivity (i.e., permeability), void ratio (i.e., porosity), and thermal conductivity of the lunar regolith. The absence of terrestrial geologic processes on the lunar surface, a lack of water, clay minerals, and organic material in lunar soils, and the narrow compositional range in terms of mineral diversity of the regolith produces a unique problem when attempting to project terrestrial soil mechanics concepts to the lunar surface. Additionally, the occurrence of agglutinates, exclusive to the lunar surface, further disassociates the physical behavior (e.g., particle crushability, flowability, etc.) of regolith from that observed in terrestrial soils.

R. Kovtun↗

[Overview of the GSFC Effort to Build Microwave Detectors for CMB Polarization]

Filled arrays of bolometers are currently being employed for use in astronomy from the far-infrared through millimeter portion of the electromagnetic spectrum. Because of the large range of wavelengths for which such detectors are applicable, the size of a pixel (p) relative to the wavelength (lambda) will vary according to the specific application of a given available technology. As the pixel size becomes comparable to a wavelength of the absorbed radiation, correlations in the radiation field at the pixel can affect the imaging properties of the system independent of the correlations introduced by the front-end optics. We study the dependence of image fidelity and induced polarization on the size of the pixel by employing a formalism in which diffraction due to the pixel boundary is treated by propagating the second-order statistical correlations of the radiation field through a model optical system. We construct simulated polarized images of square pixels for various ratios of p lambda. For the limit in which few modes are supported by the pixel (p lambda lesssiml), we find that the diffraction due to the pixel edges is non-negligible and hence must be considered along with the telescope diffraction pattern in modeling the ultimate spatial resolution of an imaging system. For the case in which the pixel is over-moded (p lambda ggl), the geometric limit is approached as expected. This technique provides a quantitative approach to understand and optimize the imaging properties of planar detectors in the few-mode limit.

Chuss, David↗

Evaluation of image quality in a Cassegrain-type telescope with an oscillating secondary mirror

A ray-trace analysis is described of aberrations and extreme rays of a Cassegrain-type telescope with a tilted secondary mirror. The work was motivated by the need to understand the factors limiting image quality and to assist in the design of secondary mirrors for three telescopes with oscillating secondary mirrors (OSM) used at Ames Research Center for high altitude infrared astronomy. The telescopes are a 31-cm-diameter Dall-Kirkham (elliptical primary, spherical secondary) flown aboard a Lear jet, a 71-cm balloon-borne Dall-Kirkham flown on the AIROscope gondola, and a 91-cm true Cassegrain (parabolic primary, hyperbolic secondary) flown aboard a C-141 jet transport. The optics for these telescopes were not designed specifically for OSM operation, but all have OSM's and all must be used with various detector configurations; therefore, a facility that evaluates the performance of a telescope for a given configuration is useful. The analytical expressions are summarized and results for the above systems are discussed. Details of the calculation and a discussion of the computer program are given in the appendices.

Erickson, E. F.↗

Context dependent prediction and category encoding for DPCM image compression

Efficient compression of image data requires the understanding of the noise characteristics of sensors as well as the redundancy expected in imagery. Herein, the techniques of Differential Pulse Code Modulation (DPCM) are reviewed and modified for information-preserving data compression. The modifications include: mapping from intensity to an equal variance space; context dependent one and two dimensional predictors; rationale for nonlinear DPCM encoding based upon an image quality model; context dependent variable length encoding of 2x2 data blocks; and feedback control for constant output rate systems. Examples are presented at compression rates between 1.3 and 2.8 bits per pixel. The need for larger block sizes, 2D context dependent predictors, and the hope for sub-bits-per-pixel compression which maintains spacial resolution (information preserving) are discussed.

Beaudet, Paul R.↗