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 397 records · Page 22
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Neutrino interaction vertex reconstruction in DUNE with Pandora deep learning

The Pandora Software Development Kit and algorithm libraries perform reconstruction of neutrino interactions in liquid argon time projection chamber detectors. Pandora is the primary event reconstruction software used at the Deep Underground Neutrino Experiment, which will operate four large-scale liquid argon time projection chambers at the far detector site in South Dakota, producing high-resolution images of charged particles emerging from neutrino interactions. While these high-resolution images provide excellent opportunities for physics, the complex topologies require sophisticated pattern recognition capabilities to interpret signals from the detectors as physically meaningful objects that form the inputs to physics analyses. A critical component is the identification of the neutrino interaction vertex. Subsequent reconstruction algorithms use this location to identify the individual primary particles and ensure they each result in a separate reconstructed particle. A new vertex-finding procedure described in this article integrates a U-ResNet neural network performing hit-level classification into the multi-algorithm approach used by Pandora to identify the neutrino interaction vertex. The machine learning solution is seamlessly integrated into a chain of pattern-recognition algorithms. The technique substantially outperforms the previous BDT-based solution, with a more than 20% increase in the efficiency of sub-1 cm vertex reconstruction across all neutrino flavours.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Facial Named Entity Recognition by Attention-Based Graph Convolutional Neural Network

In the realm of facial recognition and analysis, the ability to accurately cluster large datasets of facial images stands as a cornerstone for various applications, ranging from security surveillance to user biometric identification. This project evolves a novel approach to facial data clustering by embedding facial images into a high-dimensional vector space using an advanced embedding model trained on separate data and assumes a graph-like structure on the high-dimensional vectors. We find our method works significantly better than common shallow methods.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1973

Optical instrumentation engineering in science, technology and society; Proceedings of the Sixteenth Annual Technical Meeting, San Mateo, Calif., October 16-18, 1972

Visual tracking performance in instrumentation is discussed together with photographic pyrometry in an aeroballistic range, optical characteristics of spherical vapor bubbles in liquids, and the automatic detection and control of surface roughness by coherent diffraction patterns. Other subjects explored are related to instruments, sensors, systems, holography, and pattern recognition. Questions of data handling are also investigated, taking into account minicomputer image storage for holographic interferometry analysis, the design of a video amplifier for a 90 MHz bandwidth, and autostereoscopic screens. Individual items are announced in this issue.

Katz, Y. H.↗
NASA NTRS1982

An overview of computer vision

An overview of computer vision is provided. Image understanding and scene analysis are emphasized, and pertinent aspects of pattern recognition are treated. The basic approach to computer vision systems, the techniques utilized, applications, the current existing systems and state-of-the-art issues and research requirements, who is doing it and who is funding it, and future trends and expectations are reviewed.

Gevarter, W. B.↗
NASA NTRS1987

Improved shape-signature and matching methods for model-based robotic vision

Researchers describe new techniques for curve matching and model-based object recognition, which are based on the notion of shape-signature. The signature which researchers use is an approximation of pointwise curvature. Described here is curve matching algorithm which generalizes a previous algorithm which was developed using this signature, allowing improvement and generalization of a previous model-based object recognition scheme. The results and the experiments described relate to 2-D images. However, natural extensions to the 3-D case exist and are being developed.

Schwartz, J. T.↗
NASA NTRS1991

Hybrid solution for high-speed target acquisition and identification systems

A typical hierarchy for a general object recognition problem consists of object detection, classification and identification. This paper establishes necessary building blocks required for high-speed object recognition applications. An architecture that combines digital and optical processing, exploiting current image processing techniques for detection and classification, and optical processing hardware is described. An optical processing scheme is suggested for the identification aspect. Numerical results of each proposed concept are presented.

Udomkesmalee, Suraphol↗
NASA NTRS1995

Prototype Focal-Plane-Array Optoelectronic Image Processor

Prototype very-large-scale integrated (VLSI) planar array of optoelectronic processing elements combines speed of optical input and output with flexibility of reconfiguration (programmability) of electronic processing medium. Basic concept of processor described in "Optical-Input, Optical-Output Morphological Processor" (NPO-18174). Performs binary operations on binary (black and white) images. Each processing element corresponds to one picture element of image and located at that picture element. Includes input-plane photodetector in form of parasitic phototransistor part of processing circuit. Output of each processing circuit used to modulate one picture element in output-plane liquid-crystal display device. Intended to implement morphological processing algorithms that transform image into set of features suitable for high-level processing; e.g., recognition.

Fang, Wai-Chi↗
NASA NTRS2010

Optimization of a Multi-Stage ATR System for Small Target Identification

An Automated Target Recognition system (ATR) was developed to locate and target small object in images and videos. The data is preprocessed and sent to a grayscale optical correlator (GOC) filter to identify possible regionsof- interest (ROIs). Next, features are extracted from ROIs based on Principal Component Analysis (PCA) and sent to neural network (NN) to be classified. The features are analyzed by the NN classifier indicating if each ROI contains the desired target or not. The ATR system was found useful in identifying small boats in open sea. However, due to "noisy background," such as weather conditions, background buildings, or water wakes, some false targets are mis-classified. Feedforward backpropagation and Radial Basis neural networks are optimized for generalization of representative features to reduce false-alarm rate. The neural networks are compared for their performance in classification accuracy, classifying time, and training time.

false alarm rate↗
DOE OSTI2020

Computing for the DUNE Long-Baseline Neutrino Oscillation Experiment

This is a talk given at Computers in High Energy Physics in Adelaide, South Australia, Australia in November 2019. It is partially intended to explain the context of DUNE Computing for computing specialists. The DUNE collaboration consists of over 180 institutions from 33 countries. The experiment is in preparation now with commissioning of the first 10kT fiducial volume Liquid Argon TPC expected over the period 2025-2028 and a long data taking run with 4 modules expected from 2029 and beyond. An active prototyping program is already in place with a short test beam run with a 700T, 15,360 channel prototype of single-phase readout at the neutrino platform at CERN in late 2018 and tests of a similar sized dual-phase detector scheduled for mid-2019. The 2018 test beam run was a valuable live test of our computing model. The detector produced raw data at rates of up to ~2GB/s. These data were stored at full rate on tape at CERN and Fermilab and replicated at sites in the UK and Czech Republic. In total 1.2 PB of raw data from beam and cosmic triggers were produced and reconstructed during the six week test beam run. Baseline predictions for the full DUNE detector data, starting in the late 2020's are 30-60 PB of raw data per year. In contrast to traditional HEP computational problems, DUNE's Liquid Argon TPC data consist of simple but very large (many GB) 2D data objects which share many characteristics with astrophysical images. This presents opportunities to use advances in machine learning and pattern recognition as a frontier user of High Performance Computing facilities capable of massively parallel processing.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
NASA NTRS1974

Automated thematic mapping and change detection of ERTS-1 images

A system that inventories and updates resources must be capable of recognizing resources and their changes rapidly, using imagery acquired by a resources satellite such as ERTS-1. The conversion of ERTS images to thematic maps showing the distribution of resources is the first step in the data reduction process. To accomplish this task, the resources must be recognized from spatial and multispectral signatures. In addition, resource boundaries must be accurately established, and the data from different acquisition dates must be registered. This paper describes a system that combines multispectral and spatial pattern recognition techniques to produce thematic maps. This system has been applied to ERTS-1 MSS images, and the results obtained are discussed.

Gramenopoulos, N.↗
NASA NTRS1996

Covisualization by computational optical-sectioning microscopy of integrin and associated proteins at the cell membrane of living onion protoplasts

Using higher-resolution wide-field computational optical-sectioning fluorescence microscopy, the distribution of antigens recognized by antibodies against animal beta 1 integrin, fibronectin, and vitronectin has been visualized at the outer surface of enzymatically protoplasted onion epidermis cells and in depectinated cell wall fragments. On the protoplast all three antigens are colocalized in an array of small spots, as seen in raw images, in Gaussian filtered images, and in images restored by two different algorithms. Fibronectin and vitronectin but not beta 1 integrin antigenicities colocalize as puncta in comparably prepared and processed images of the wall fragments. Several control visualizations suggest considerable specifity of antibody recognition. Affinity purification of onion cell extract with the same anti-integrin used for visualization has yielded protein that separates in SDS-PAGE into two bands of about 105-110 and 115-125 kDa. These bands are again recognized by the visualization antibody, which was raised against the extracellular domain of chicken beta 1 integrin, and are also recognized by an antibody against the intracellular domain of chicken beta 1 integrin. Because beta 1 integrin is a key protein in numerous animal adhesion sites, it appears that the punctate distribution of this protein in the cell membranes of onion epidermis represents the adhesion sites long known to occur in cells of this tissue. Because vitronectin and fibronection are matrix proteins that bind to integrin in animals, the punctate occurrence of antigenically similar proteins both in the wall (matrix) and on enzymatically prepared protoplasts reinforces the concept that onion cells have adhesion sites with some similarity to certain kinds of adhesion sites in animals.

NASA Discipline Plant Biology↗
NASA NTRS1992

Automated Recognition Of Melt Lines In Phase-Change Paints

Technique developed to reduce data quickly and objectively in images of phase-change paints. Involves use of three-tube color-camera system. Camera modified so separate red (R), green (G), and blue (B) video outputs contain complete framing information recorded and digitized by use of standard video format. Proven very simple and repeatable. Used wherever rapid and accurate remote temperature-profile measurements desired; for example, for test objects in inaccessible, toxic, or other hostile areas.

Buck, Gregory M.↗
NASA NTRS1970

Multispectral analysis of limestone, dolomite, and granite, Mill Creek, Oklahoma

Spectral reflectance and thermal emission data were collected at the Mill Creek, Oklahoma test site during NASA missions 132 and 133 in June 1970. The data were collected by three aircraft flown several times during the diurnal cycle at altitudes of 150 to 17,000 m above mean terrain. Reflectance of the main rock types (limestone, dolomite, and granite) was determined from the data collected using a 12-channel multispectral scanner during mission 133 and from thermal infrared images recorded during mission 132 on an RS-7 scanner from 17,000 m above terrain. A preliminary rock recognition map was generated automatically using data collected from 900 m above terrain. The discrimination provided by the map is reasonably accurate. Misidentification occurred in areas of unusually high dolomite reflectivity. High altitude thermal infrared (10 to 12 micrometers) images show regional folds and faults distinguished by the presence of thermally contrasting materials. Linear and curvilinear structural features two to three times smaller than the nominal 17 m resolution could be detected.

Rowan, L. C.↗
NASA NTRS2004

Software and Algorithms for Biomedical Image Data Processing and Visualization

A new software equipped with novel image processing algorithms and graphical-user-interface (GUI) tools has been designed for automated analysis and processing of large amounts of biomedical image data. The software, called PlaqTrak, has been specifically used for analysis of plaque on teeth of patients. New algorithms have been developed and implemented to segment teeth of interest from surrounding gum, and a real-time image-based morphing procedure is used to automatically overlay a grid onto each segmented tooth. Pattern recognition methods are used to classify plaque from surrounding gum and enamel, while ignoring glare effects due to the reflection of camera light and ambient light from enamel regions. The PlaqTrak system integrates these components into a single software suite with an easy-to-use GUI (see Figure 1) that allows users to do an end-to-end run of a patient s record, including tooth segmentation of all teeth, grid morphing of each segmented tooth, and plaque classification of each tooth image. The automated and accurate processing of the captured images to segment each tooth [see Figure 2(a)] and then detect plaque on a tooth-by-tooth basis is a critical component of the PlaqTrak system to do clinical trials and analysis with minimal human intervention. These features offer distinct advantages over other competing systems that analyze groups of teeth or synthetic teeth. PlaqTrak divides each segmented tooth into eight regions using an advanced graphics morphing procedure [see results on a chipped tooth in Figure 2(b)], and a pattern recognition classifier is then used to locate plaque [red regions in Figure 2(d)] and enamel regions. The morphing allows analysis within regions of teeth, thereby facilitating detailed statistical analysis such as the amount of plaque present on the biting surfaces on teeth. This software system is applicable to a host of biomedical applications, such as cell analysis and life detection, or robotic applications, such as product inspection or assembly of parts in space and industry.

Talukder, Ashit↗
NASA NTRS2006

Detection of Water Hazards for Autonomous Robotic Vehicles

Four methods of detection of bodies of water are under development as means to enable autonomous robotic ground vehicles to avoid water hazards when traversing off-road terrain. The methods involve processing of digitized outputs of optoelectronic sensors aboard the vehicles. It is planned to implement these methods in hardware and software that would operate in conjunction with the hardware and software for navigation and for avoidance of solid terrain obstacles and hazards. The first method, intended for use during the day, is based on the observation that, under most off-road conditions, reflections of sky from water are easily discriminated from the adjacent terrain by their color and brightness, regardless of the weather and of the state of surface waves on the water. Accordingly, this method involves collection of color imagery by a video camera and processing of the image data by an algorithm that classifies each pixel as soil, water, or vegetation according to its color and brightness values (see figure). Among the issues that arise is the fact that in the presence of reflections of objects on the opposite shore, it is difficult to distinguish water by color and brightness alone. Another issue is that once a body of water has been identified by means of color and brightness, its boundary must be mapped for use in navigation. Techniques for addressing these issues are under investigation. The second method, which is not limited by time of day, is based on the observation that ladar returns from bodies of water are usually too weak to be detected. In this method, ladar scans of the terrain are analyzed for returns and the absence thereof. In appropriate regions, the presence of water can be inferred from the absence of returns. Under some conditions in which reflections from the bottom are detectable, ladar returns could, in principle, be used to determine depth. The third method involves the recognition of bodies of water as dark areas in short-wavelength infrared (SWIR) images. This method is based on the fact, well known among experts in remote sensing, that water bodies of any appreciable depth appear very dark in near-infrared, overhead imagery. Even under a thick layer of marine fog, SWIR illumination is present. Hence, this method may work even in the presence of clouds, though it is unlikely to work at night. Snow and ice also exhibit very strong absorption at wavelengths greater than about 1.4 m. Hence, the wavelength range of about 1.5 to 1.6 m might be useable in this method for recognizing water, snow, and ice. One notable drawback of this method is that useful look-ahead distance could be limited by surface reflections. The fourth method, intended for use at night, involves the contrast between water and terrain in thermal-infrared (medium-wavelength infrared) imagery. This method is based on the fact that at night, water is usually warmer than the adjacent terrain. Look-ahead distance could be limited in this method because, for reasons not yet fully understood, water appears to darken in the thermal infrared with increasing distance.

Matthes, Larry↗
NASA NTRS1975

A boundary finding algorithm and its applications

An algorithm for locating gray level and/or texture edges in digitized pictures is presented. The algorithm is based on the concept of hypothesis testing. The digitized picture is first subdivided into subsets of picture elements, e.g., 2 x 2 arrays. The algorithm then compares the first- and second-order statistics of adjacent subsets; adjacent subsets having similar first- and/or second-order statistics are merged into blobs. By continuing this process, the entire picture is segmented into blobs such that the picture elements within each blob have similar characteristics. The boundaries between the blobs comprise the boundaries. The algorithm always generates closed boundaries. The algorithm was developed for multispectral imagery of the earth's surface. Application of this algorithm to various image processing techniques such as efficient coding, information extraction (terrain classification), and pattern recognition (feature selection) are included.

Gupta, J. N.↗
NASA NTRS1984

Interlobate comparison of glacial-depositional style as evidenced by small-relief glacial landscape features in Illinois, Indiana, and Ohio, utilizing SIR-B

The objectives and expected results of an investigation of the use of the Shuttle Imaging Radar-B (SIR-B) as a basic tool in the recognition and mapping of glacial landforms are discussed. The main goals are: (1) to evaluate the ability of SIR-B to delineate varying sizes, shapes, and relief of surface forms; (2) to compare and contrast SIR-B imagery with selected Seasat SAR imagery; (3) to utilize SIR-B imagery synergistically with available SEASAT SAR, LANDSAT RBV, and other imagery sources to identify and map suites of glacial landforms; and (4) eventually to interpret the suites in terms of ice dynamics and conditions of deglaciation, to relate them to the stratigraphic record, and to evaluate interactions of the major lobes and sublobes.

Johnson, W. H.↗