SEARCH · Engineering Papers
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.
The solar chromosphere consists of three classes which contribute differently to ultraviolet radiation reaching the earth. We describe a data set of solar images, means of segmenting the images into the constituent classes, and novel high-level representation for compact objects based on a triangulation spatial 'membership function'.
Bartron Medical Imaging LLC (BMI), of New Haven, Connecticut, gained a nonexclusive license from Goddard Space Flight Center to use the RHSEG software in medical imaging. To manage image data, BMI then licensed two pattern-matching software programs from NASA's Jet Propulsion Laboratory that were used in image analysis and three data-mining and edge-detection programs from Kennedy Space Center. More recently, BMI made NASA history by being the first company to partner with the Space Agency through a Cooperative Research and Development Agreement to develop a 3-D version of RHSEG. With U.S. Food and Drug Administration clearance, BMI will sell its Med-Seg imaging system with the 2-D version of the RHSEG software to analyze medical imagery from CAT and PET scans, MRI, ultrasound, digitized X-rays, digitized mammographies, dental X-rays, soft tissue analyses, moving object analyses, and soft-tissue slides such as Pap smears for the diagnoses and management of diseases. Extending the software's capabilities to three dimensions will eventually enable production of pixel-level views of a tumor or lesion, early identification of plaque build-up in arteries, and identification of density levels of microcalcification in mammographies.
Fully two-dimensional gamma-ray imaging with simultaneous high-resolution spectroscopy has been demonstrated using an externally segmented germanium sensor. The system employs a single high-purity coaxial detector with its outer electrode segmented into 5 distinct charge collection regions and a lead coded aperture with a uniformly redundant array (URA) pattern. A series of one-dimensional responses was collected around 511 keV while the system was rotated in steps through 180 degrees. A non-negative, linear least-squares algorithm was then employed to reconstruct a 2-dimensional image. Corrections for multiple scattering in the detector, and the finite distance of source and detector are made in the reconstruction process.
Compressing technique calculates activity estimator for each segment of image line. Estimator is used in conjunction with allowable bits per line, N, to determine number of bits necessary to code each segment and which segments can tolerate truncation. Preprocessed line data are then passed to adaptive variable-length coder, which selects optimum transmission code. Method increases capacity of broadcast and cable television transmissions and helps reduce size of storage medium for video and digital audio recordings.
This project, performed in support of the NASA GRC Space Academy summer program, sought to develop an open-source workflow methodology that segmented medical image data, created a 3D model from the segmented data, and prepared the model for finite-element analysis. In an initial step, a technological survey evaluated the performance of various existing open-source software that claim to perform these tasks. However, the survey concluded that no single software exhibited the wide array of functionality required for the potential NASA application in the area of bone, muscle and bio fluidic studies. As a result, development of a series of Python scripts provided the bridging mechanism to address the shortcomings of the available open source tools. The implementation of the VTK library provided the most quick and effective means of segmenting regions of interest from the medical images; it allowed for the export of a 3D model by using the marching cubes algorithm to build a surface mesh. To facilitate the development of the model domain from this extracted information required a surface mesh to be processed in the open-source software packages Blender and Gmsh. The Preview program of the FEBio suite proved to be sufficient for volume filling the model with an unstructured mesh and preparing boundaries specifications for finite element analysis. To fully allow FEM modeling, an in house developed Python script allowed assignment of material properties on an element by element basis by performing a weighted interpolation of voxel intensity of the parent medical image correlated to published information of image intensity to material properties, such as ash density. A graphical user interface combined the Python scripts and other software into a user friendly interface. The work using Python scripts provides a potential alternative to expensive commercial software and inadequate, limited open-source freeware programs for the creation of 3D computational models. More work will be needed to validate this approach in creating finite-element models.
Texture is an important image characteristic. A variety of spatial domain techniques were proposed for extracting and utilizing textural features for segmenting and classifying images. for the most part, these spatial domain techniques are ad hos in nature. A markov random field model for image texture is discussed. A frequency domain description of image texture is derived in terms of the power spectral density. This model is used for designing optimum frequency domain filters for enhancing, restoring and segmenting images based on their textural properties.
Evaluating the mechanical response of fiber-reinforced composites can be extremely time-consuming and expensive. Machine learning (ML) techniques offer a means for faster predictions via models trained on existing input–output pairs and have exhibited success in composite research. This paper explores a fully convolutional neural network modified from StressNet, which was originally used for linear elastic materials, and extended here for a non-linear finite element (FE) simulation to predict the stress field in 2D slices of segmented tomography images of a fiber-reinforced polymer specimen. The network was trained and evaluated on data generated from the FE simulations of the exact microstructure. The testing results show that the trained network accurately captures the characteristics of the stress distribution, especially on fibers, solely from the segmented microstructure images. The trained model can make predictions within seconds in a single forward pass on an ordinary laptop, given the input microstructure, compared to 92.5 h to run the full FE simulation on a high-performance computing cluster. These results show promise in using ML techniques to conduct fast structural analysis for fiber-reinforced composites and suggest a corollary that the trained model can be used to identify the location of potential damage sites in fiber-reinforced polymers.
The recent introduction of deep learning methods for image processing has greatly advanced the characterization of materials using three-dimensional (3D) X-ray imaging techniques. However, deep learning models often have difficulty performing consistently across images owing to unavoidable variations in imaging conditions, which create inconsistencies even for the same material. As a result, networks must frequently be retrained for new datasets, limiting their applicability and generalization. Thus, it is critical to reduce the variations between images to enable a single model to process multiple datasets. Herein, we introduce P3T-Net, a pseudo-3D domain transfer network that transfers diverse 3D images into a uniform domain before processing using deep learning models. Remarkably, P3T-Net enables the reuse of previously trained networks for processing new images and considerably reduces the computational cost of transferring 3D images across domains. These unique capabilities were demonstrated in the following scenarios: (i) image enhancement of fast scans for geological rock and hydrogen fuel cells, (ii) enhancement of images to match the quality of multi-source imaging for lithium-ion batteries, (iii) accurate segmentation of images captured under different conditions, and (iv) tera-scale 3D transfer (10 11 voxels) on a single GPU. Overall, the proposed approach addresses cross-domain inconsistencies across various materials and conditions, thereby enabling more robust and generalizable deep learning solutions for a wide range of material imaging tasks.
Explore the source record for details and available documents.
The intelligent visual system detects anomalies or defects in real time under normal lighting operating conditions. The application is basically a learning machine that integrates fuzzy logic (FL), artificial neural network (ANN), and generic algorithm (GA) schemes to process the image, run the learning process, and finally detect the anomalies or defects. The system acquires the image, performs segmentation to separate the object being tested from the background, preprocesses the image using fuzzy reasoning, performs the final segmentation using fuzzy reasoning techniques to retrieve regions with potential anomalies or defects, and finally retrieves them using a learning model built via ANN and GA techniques. FL provides a powerful framework for knowledge representation and overcomes uncertainty and vagueness typically found in image analysis. ANN provides learning capabilities, and GA leads to robust learning results. An application prototype currently runs on a regular PC under Windows NT, and preliminary work has been performed to build an embedded version with multiple image processors. The application prototype is being tested at the Kennedy Space Center (KSC), Florida, to visually detect anomalies along slide basket cables utilized by the astronauts to evacuate the NASA Shuttle launch pad in an emergency. The potential applications of this anomaly detection system in an open environment are quite wide. Another current, potentially viable application at NASA is in detecting anomalies of the NASA Space Shuttle Orbiter's radiator panels.
MLExchange is a machine learning (ML) operations platform providing web user-interfaces (UIs) for data visualization and analysis pipelines at synchrotron facilities. Among these UIs is the segmentation app which helps synchrotron users utilize ML algorithms to automatically segment high-resolution scientific images with minimal manual annotation effort. In this work, we share code optimizations that significantly speed up the segmentation inference workflow of large data in short time. By optimizing the sequence of CPU-GPU data transfers and introducing CPU parallelization to key operations, we improve the per-device, per-image frame computational efficiency and observe close to 3×$$\times$$ speedup over the original segmentation inference workflow run time when utilizing a single GPU. Further adaptations enabling multi-GPU inference yield more than 40×$$\times$$ speedup with 100 GPUs compared to the optimized single GPU inference workflow. This acceleration of the segmentation inference workflow will provide MLExchange users with easy access to segmentation results with little wait time.
Artificial neural networks that would utilize the cascade error projection (CEP) algorithm have been proposed as means of autonomous, real-time, adaptive color segmentation of images that change with time. In the original intended application, such a neural network would be used to analyze digitized color video images of terrain on a remote planet as viewed from an uninhabited spacecraft approaching the planet. During descent toward the surface of the planet, information on the segmentation of the images into differently colored areas would be updated adaptively in real time to capture changes in contrast, brightness, and resolution, all in an effort to identify a safe and scientifically productive landing site and provide control feedback to steer the spacecraft toward that site. Potential terrestrial applications include monitoring images of crops to detect insect invasions and monitoring of buildings and other facilities to detect intruders. The CEP algorithm is reliable and is well suited to implementation in very-large-scale integrated (VLSI) circuitry. It was chosen over other neural-network learning algorithms because it is better suited to realtime learning: It provides a self-evolving neural-network structure, requires fewer iterations to converge and is more tolerant to low resolution (that is, fewer bits) in the quantization of neural-network synaptic weights. Consequently, a CEP neural network learns relatively quickly, and the circuitry needed to implement it is relatively simple. Like other neural networks, a CEP neural network includes an input layer, hidden units, and output units (see figure). As in other neural networks, a CEP network is presented with a succession of input training patterns, giving rise to a set of outputs that are compared with the desired outputs. Also as in other neural networks, the synaptic weights are updated iteratively in an effort to bring the outputs closer to target values. A distinctive feature of the CEP neural network and algorithm is that each update of synaptic weights takes place in conjunction with the addition of another hidden unit, which then remains in place as still other hidden units are added on subsequent iterations. For a given training pattern, the synaptic weight between (1) the inputs and the previously added hidden units and (2) the newly added hidden unit is updated by an amount proportional to the partial derivative of a quadratic error function with respect to the synaptic weight. The synaptic weight between the newly added hidden unit and each output unit is given by a more complex function that involves the errors between the outputs and their target values, the transfer functions (hyperbolic tangents) of the neural units, and the derivatives of the transfer functions.
A simple knowledge-based approach to the recognition of objects in man-made scenes is being developed. Specifically, the system under development is a proposed enhancement to a robot arm for use in the space station laboratory module. The system will take a request from a user to find a specific object, and locate that object by using its camera input and information from a knowledge base describing the scene layout and attributes of the object types included in the scene. In order to use realistic test images in developing the system, researchers are using photographs of actual NASA simulator panels, which provide similar types of scenes to those expected in the space station environment. Figure 1 shows one of these photographs. In traditional approaches to image analysis, the image is transformed step by step into a symbolic representation of the scene. Often the first steps of the transformation are done without any reference to knowledge of the scene or objects. Segmentation of an image into regions generally produces a counterintuitive result in which regions do not correspond to objects in the image. After segmentation, a merging procedure attempts to group regions into meaningful units that will more nearly correspond to objects. Here, researchers avoid segmenting the image as a whole, and instead use a knowledge-directed approach to locate objects in the scene. The knowledge-based approach to scene analysis is described and the categories of knowledge used in the system are discussed.
We have superpolished a diamond-turned aluminum mandrel (coated with electroless Ni) to an axial roughness of 6.34 nm rms. The mandrel is made to the Astro-E secondary mirror design for the 81st shell. Precision metrology at 100 mm to submicron scales has established the power spectral density of the mandrel and ultralightweight gold coated replicated segments. Predicted image quality of a set of optimally aligned replicated segments of this and a matching primary is substantially improved as compared to the flight mirrors for Astro-E. This approach using metal mandrels, superpolishing, and replicated ultralightweight foil mirrors, may represent a cost-effective approach to meeting the 15 arcsec half-energy width and weight requirements for the Constellation-X mission. Descriptions of the polishing apparatus, the precision metrology instruments, and the surface data analysis are presented. The general methods described are applicable to precision optics for both normal incidence and grazing incidence optics.
We have superpolished a diamond-turned aluminum mandrel (coated with electroless Ni) to an axial roughness of 0.34 nm rms. The mandrel is made to the Astro-E secondary mirror design for the 81st shell. Precision metrology at 100 mm to submicron scales has established the power spectral density of the mandrel and ultralightweight gold coated replicated segments. Predicted image quality of a set of optimally aligned replicated segments of this and a matching primary is substantially improved as compared to the flight mirrors for Astro-E. This approach using metal mandrels, superpolishing, and replicated ultralightweight foil mirrors, may represent a cost-effective approach to meeting the 15 arcsec half-energy width and weight requirements for the Constellation-X mission. Descriptions of the polishing apparatus, the precision metrology instruments, and the surface data analysis are presented. The general methods described are applicable to precision optics for both normal incidence and grazing incidence optics.
We have superpolished a diamond-turned aluminum mandrel (coated with electroless Ni) to an axial roughness of 0.34 nm rms. The mandrel is made to the Astro-E secondary mirror design for the 81st shell. Precision metrology at 100 mm to submicron scales has established the power spectral density of the mandrel and ultralightweight gold coated replicated segments. Predicted image quality of a set of optimally aligned replicated segments of this and a matching primary is substantially improved as compared to the flight mirrors for Astro-E. This approach using metal mandrels, superpolishing, and replicated ultralightweight foil mirrors, may represent a cost-effective approach to meeting the 15 arcsec half-energy width and weight requirements for the Constellation-X mission. Descriptions of the polishing apparatus, the precision metrology instruments, and the surface data analysis are presented. The general methods described are applicable to precision optics for both normal incidence and grazing incidence optics.
In many applications one would like to use information from both color and texture features in order to segment an image. We propose a novel technique to combine "soft" segmentations computed for two or more features independently. Our algorithm merges models according to a mean entropy criterion, and allows to choose the appropriate number of classes for the final grouping. This technique also allows to improve the quality of supervised classification based on one feature (e.g. color) by merging information from unsupervised segmentation based on another feature (e.g., texture.)
An unsupervised clustering algorithm is used to segment multifrequency polarimetric radar data from the NASA/JPL airborne SAR (synthetic aperture radar). Twenty-two parameters are evaluated for their discriminatory capability for each pixel of an image. A clustering analysis is then performed using different subsets of these parameters. This analysis relies on data taken as part of an intensive field experiment during the summer of 1988 in the vicinity of the Pisgah lava flow in the Mojave Desert in southern California. As part of the experiment, extensive ground truth was acquired, including dielectric constant and topography measurements. Segmentation results show good agreement with these measurements.