Three Dimensional Urban Characterization by IFSAR Measurements
In this paper a machine vision approach is applied to IFSAR data to extract the most relevant built structures in a dense urban environment.
IFSAR↗
SEARCH · Search NASA
Results for “Machine vision”
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.
CMOS Active Pixel Image Sensor
A new CMOS active pixel image sensor is reported. The sensor uses a 2.0mue double-poly, double-metal foundry CMOS process and is realized as a 28 x 28 array of 40 mue x 40 mue pixels. The Sensor features TTL compatible voltages, low noise and large dynamic range, and will be useful in machine vision and smart sensor applications.
CMOS↗
Detection and Tracking of Aircraft in the Far-Field from Small Unmanned Aerial Systems
Onboard far-field aircraft detection is needed for safe non-cooperative traffic mitigation in autonomous small Unmanned Aerial System (sUAS) operations. Machine vision systems, based on standard optics and visible light detectors, possess the ideal size, weight, and power (SWaP) requirements for sUAS. This work presents the design and analysis of a novel aircraft detection and tracking pipeline based on optical sensing alone. Key contributions of the work include a refined range inequality model based on sensing and detection with FAA well-clear separation assurance distances between aircraft in mind, a detector fusion method to maximize the benefit of two image detectors, and a comparative analysis of Linear Kalman-filtering and Extended Kalman-filtering to seek optimal tracking performance. The pipeline is evaluated offline against multiple intruder platforms, using two types of flight encounters: multirotor sUAS vs. fixed-wing sUAS and multirotor sUAS vs. general aviation(GA)plane. Analysis is restricted to the rate-limiting head-on and departing collision volume cases vertically separated for safety. Results indicate that it is feasible to use the proposed optical spatial-temporal tracking algorithm to provide adequate alerting time to prevent penetration of well-clear separation volumes for both sUAS and GA aircraft.
Unmanned Aerial System↗
Towards the Development of a Multi-Agent Cognitive Networking System for the Lunar Environment
This paper details the development of a multi-agent cognitive system intended to optimize networking performance in the lunar environment. NASA’s current concept of the future of lunar communication, LunaNet [1], outlines a complex network of networks. Challenges such as scalability, interoperability and reliability must first be addressed to successfully fulfill this vision. Machine intelligence can greatly reduce the reliance on human operators and enable efficient operations for tasks such as scheduling and network management. The application of machine learning, artificial intelligence, and other automated decision-making techniques can be used to allow network nodes to intelligently sense and adapt to changes in the environment such as link disruptions, new nodes joining the network, and support for a diverse range of protocols. Cognitive networking seeks to evolve these technologies into an autonomous system with improved science data return, reliability, and scalability. In this paper, we study three main areas a means to further develop cognitive networking capabilities: networking and flight software development, analysis of wireless data for modeling and simulation, and development of algorithms for a multi-agent system.
Cognitive Networking↗
Towards the Development of a Multi-Agent Cognitive Networking System for the Lunar Environment
This paper details the development of a multi-agent cognitive system intended to optimize networking performance in the lunar environment. NASA’s current concept of the future of lunar communication, LunaNet, outlines a complex network of networks. Challenges such as scalability, interoperability and reliability must first be addressed to successfully fulfill this vision. Machine intelligence can greatly reduce the reliance on human operators and enable efficient operations for tasks such as scheduling and network management. The application of machine learning, artificial intelligence, and other automated decision-making techniques can be used to allow network nodes to intelligently sense and adapt to changes in the environment such as link disruptions, new nodes joining the network, and support for a diverse range of protocols. Cognitive networking seeks to evolve these technologies into an autonomous system with improved science data return, reliability, and scalability. In this paper, we study three main areas a means to further develop cognitive networking capabilities: networking and flight software development, analysis of wireless data for modeling and simulation, and development of algorithms for a multi-agent system.
Rachel Dudukovich↗
Towards the Development of a Multi-Agent Cognitive Networking System for the Lunar Environment
This paper details the development of a multi-agent cognitive system intended to optimize networking performance in the lunar environment. One concept of the future of lunar communication, LunaNet, outlines a complex network of networks. Challenges such as scalability, interoperability, and reliability must first be addressed to successfully fulfill this vision. Machine intelligence can greatly reduce the reliance on human operators and enable efficient operations for tasks such as scheduling and network management. Machine learning, artificial intelligence, and other automated decision-making techniques can be used to allow network nodes to intelligently sense and adapt to changes in the environment such as link disruptions, new nodes joining the network, and support for a diverse range of protocols. Cognitive networking seeks to evolve these technologies into an autonomous system with improved science data return, reliability, and scalability. In this paper, we study four main areas as a means to further develop cognitive networking capabilities: networking protocol development, analysis of wireless data for modeling and simulation, development of algorithms for a multi-agent system, and spectrum sensing technology.
cognitive networking↗
AI4MARS: A Dataset for Terrain-Aware Autonomy on Mars
Deep learning has quickly become a necessity for selfdriving vehicles on Earth. In contrast, the self-driving vehicles on Mars, including NASA’s latest rover, Perseverance, which is planned to land on Mars in February 2021, are still driven by classical machine vision systems. Deep learning capabilities, such as semantic segmentation and object recognition, would substantially benefit the safety and productivity of ongoing and future missions to the red planet. To this end, we created the first large-scale dataset, AI4Mars, for training and validating terrain classification models for Mars, consisting of ~326K semantic segmentation full image labels on 35K images from Curiosity, Opportunity, and Spirit rovers, collected through crowdsourcing. Each image was labeled by ~10 people to ensure greater quality and agreement of the crowdsourced labels. It also includes ~1.5K validation labels annotated by the rover planners and scientists from NASA’s MSL (Mars Science Laboratory) mission, which operates the Curiosity rover, and MER (Mars Exploration Rovers) mission, which operated the Spirit and Opportunity rovers. We trained a DeepLabv3 model on the AI4Mars training dataset and achieved over 96% overall classification accuracy on the test set. The dataset is made publicly available.1
Ono, Hiro↗
Benchmarking and Testing of Qualcomm Snapdragon System-on-Chip for JPL Space Applications and Missions
As some space missions become more challenging due to new environments, greater distances, or more limited size, weight, and power (SWaP) constraints, spacecraft avionics must adapt to allow the spacecraft to be more autonomous and agile---eliminating the Spacecraft-Earth-Spacecraft feedback loop whenever possible. Prime examples of such missions include Aerobots (such as Ingenuity with extremely low SWaP constraints and demanding signal/image processing during flight) and landers in possibly hostile environments (such as a Europa lander mission, with limited communication capacity, high latency, and constrained power budget). To address these challenges, JPL worked with Qualcomm to demonstrate the use of their Snapdragon 801 system-on-chip (SoC) onboard the Ingenuity Helicopter on Mars. The Qualcomm Snapdragon SoC contains various subsystems, including an ARM cluster, a Graphics processing unit, a Digital Signal Processing subsystem, a Neural Processing Engine, Image Signal Processing subsystem, among others. Since the success of Ingenuity, JPL is continuing to work with Qualcomm to address other applications of the Snapdragon SoC technology. This includes the deployment of two 855 Snapdragon development boards onboard the International Space Station (ISS) for successful in-situ benchmarking of applications in space (beyond those tested on Ingenuity). In this paper, we will examine the performance of various applications that have been identified to benefit from greater onboard computational capability. These applications include (among others): machine vision algorithms that are expected to be critical in autonomous entry-descent-and-landing scenarios and real-time Aerobot flight navigation; Hyperspectral compression algorithms; Synthetic Aperture Radar Processing along with various instrument processing algorithms. We discuss how the infusion of Qualcomm's Snapdragon SoC is capable of enabling missions that may not have been able to achieve their goals with traditional flight computing. In addition, we also show that for some algorithms, the software implementation on the Snapdragon SoC outperforms traditional FPGA implementations.
Cretu, Vlad↗
Ionizing-radiation beam monitoring system
A transmissive ionizing-radiation beam monitoring system includes an enclosure structure with at least one ultra-thin window to an incident ionizing-radiation beam, where the ultra-thin window is highly transmissive to ionizing-radiation. Embodiments include at least one thin or ultra-thin scintillator within the enclosure structure that is substantially directly in an incident ionizing-radiation beam path and transmissive to the incident radiation beam, and at least one ultraviolet (UV) illumination source within the enclosure structure facing the scintillator. Embodiments include at least one machine vision camera within the enclosure structure located out of an incident ionizing-radiation beam path and including a camera body and lens having a projection of its optical axis oriented at an angle of incidence of 45±35 degrees to a surface of the scintillator.
Friedman, Peter S.↗
Ionizing-radiation beamline monitoring system
Embodiments are directed generally to an ionizing-radiation beamline monitoring system that includes a vacuum chamber structure with vacuum compatible flanges through which an incident ionizing-radiation beam enters the monitoring system. Embodiments further include at least one scintillator within the vacuum chamber structure that can be at least partially translated in the ionizing-radiation beam while oriented at an angle greater than 10 degrees to a normal of the incident ionizing-radiation beam, a machine vision camera coupled to a light-tight structure at atmospheric/ambient pressure that is attached to the vacuum chamber structure by a flange attached to a vacuum-tight viewport window with the camera and lens optical axis oriented at an angle of less than 80 degrees with respect to a normal of the scintillator, and at least one ultraviolet (“UV”) illumination source facing the scintillator in the ionizing-radiation beam for monitoring a scintillator stability comprising scintillator radiation damage.
Friedman, Peter S.↗
Method and system for automated plant surveillance and manipulation
A method and a system for automated plant surveillance and manipulation are provided. Pursuant to the method and the system, images of target plants are obtained through a machine vision system having multiple cameras. The obtained images of the target plants are processed to determine tissue candidates of the target plants and to determine a position and an orientation of each tissue candidate. A tool is manipulated, based on the position and the orientation of each tissue candidate, to excise each tissue candidate to obtain tissue samples. The tissue samples are transported for subsequently manipulation including live processing of the tissue samples or destructive processing of the tissue samples.
Kalluri, Udaya C.↗
Ionizing-radiation beamline monitoring system
Embodiments are directed generally to an ionizing-radiation beamline monitoring system that includes a vacuum chamber structure with vacuum compatible flanges through which an incident ionizing-radiation beam enters the monitoring system. Embodiments further include at least one scintillator within the vacuum chamber structure that can be at least partially translated in the ionizing-radiation beam while oriented at an angle greater than 10 degrees to a normal of the incident ionizing-radiation beam, a machine vision camera coupled to a light-tight structure at atmospheric/ambient pressure that is attached to the vacuum chamber structure by a flange attached to a vacuum-tight viewport window with the camera and lens optical axis oriented at an angle of less than 80 degrees with respect to a normal of the scintillator, and at least one ultraviolet (“UV”) illumination source facing the scintillator in the ionizing-radiation beam for monitoring a scintillator stability comprising scintillator radiation damage.
Friedman, Peter S.↗
Collaborative Robots vs. Master/Slaves - 19522
This paper will discuss the feasibility, application, and economy of applying collaborative robots (co-bots) in lieu of traditional through-the-wall tele-operated manipulator systems, traditionally used within hazardous (radioactive) environments. Both fixed facility and mobile applications will be discussed. Both single- and dual-armed co-bot systems will be employed. The six-axis, plus gripper, single-armed co-bots will provide for machine vision enhanced pick-n-place applications within shielded rooms and also include force-reflecting (haptic) teleoperations controls for non-routine materials-handling operations. The dual-armed system will be mounted to a motorized stand with motorized adjustment both vertically and rotationally (around its base). To complement these co-bots, both two-dimensional and three-dimensional vision technologies will be implemented. The vision technologies, in these applications, are used in lieu of traditional radiation shielding windows. two-dimensional video systems will provide for large area viewing, incorporating pan, tilt, and zoom capabilities. Three-dimensional video systems will be mounted to the dual-armed robot stand and provide viewing, including depth-of-perception, to the operator, performing material handling and other tasks, from an operator workstation located outside the hazardous environment. The unique workstation used with the dual-armed systems will also be described. (authors)
12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
A new probe concept for internal pipework inspection - 241
The interior visual inspection of nuclear pipework is a critical inspection activity required to ensure the continued safe, reliable operation of plant and thus avoid costly outages. Typically, the video output from a manually deployed probe is viewed by an operator online with the task of identifying and estimating the location of surface defects such as cracks, corrosion and pitting. However, it is very difficult to estimate the nature and spatial extent of defects from the often disorientating small field of view video of a relatively large structure. This work describes a new visual inspection system incorporating a wide field of view machine vision camera and additional sensors designed for inspecting 3 - 6 inch diameter pipes. The output of the system is a photorealistic model of the internal surface of the pipework. The generation of this model relies upon a core component of the system in the form of image feature extraction which estimates camera location. This paper considers the accuracy of this estimation as a function of the number and configuration of the extracted image features. (authors)
42 ENGINEERING↗
Using Fiducial Markers for Pose Estimation of an OSWEC in a Wave Tank: Preprint
In this study, we consider a novel method of sensing the motion of a wave energy converter during testing in a wave flume under the influence of incoming waves. The wave energy converter considered in our research is an oscillating surge wave energy converter, which is a hinged paddle that responds to incoming waves. Motion sensing is normally done with inertial sensors, which can hinder the motion due to suspended cables that carry power and transmit signals. Our proposed method is contactless and can be implemented economically. A camera is used to record different marker patterns affixed to the moving paddle and the motion deduced by pose estimation algorithms. Fiducial markers are commonly used for robot localization and in augmented reality. There are many types of fiducial markers, including ArUco-type markers which are accurate, fast and robust. The system consists of markers attached to the paddle element and recorded using a machine vision camera. A pose estimation algorithm is then applied to the detected markers to estimate the tilt of the paddle. In this work, we examine the challenges of image acquisition and calibration for underwater targets, compare the motion obtained by this new system with a calibrated tilt sensor and identify areas where the new system may be superior.
computer vision↗
Thermographic phosphor digital image correlation
The invention advances the ratio method of phosphor thermography using two machine vision cameras for full-field temperature measurements of a solid surface. As an example, algorithms from digital image correlation (DIC) can be used to determine the stereoscopic imaging system intrinsic and extrinsic parameters, and accurately register material points on the sample to subpixel locations in each image with 0.07 px or better accuracy. A phosphor calibration sample fabricated using aerosol deposition can be used for in situ determination of the temperature-versus-intensity ratio relationship. The full calibration methodology and several improvements on two-color phosphor thermography open the door for full-field temperature measurements in dynamic tests with deforming test specimens. In particular, thermographic phosphors can be combined with stereo digital image correlation in a novel diagnostic, TP+DIC, to measure time-resolved full-field surface strains and temperatures simultaneously, thereby enabling testing in combined thermo-mechanical environments.
Winters, Caroline↗
Process Image Analysis using Big Data, Machine Learning, and Computer Vision
The development of algorithms for machine learning and data analysis for the 3013 MIS corrosion surveillance program is a collaborative effort by SRNL, USC and GT. For corrosion detection, LCM image data is extracted from large binary files, with software written to convert the data to physical attributes (i.e. height, color and grayscale values; all as functions of a location in a plane projection). The user interface for the software permits selective downloading of binary data and interrogation of attributes. User input thresholds are used to flag attributes of interest. Machine learning algorithms, developed for this application, are used to determine whether the features are the result of corrosion. To address the fundamental mechanisms of corrosion, machine learning algorithms are being developed to derive interatomic potential force-fields from ab-initio DFT calculations. The goal is to apply molecular modeling on a large enough scale to guide the design of resistant materials.
12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
Rapid Assessment and Optimization of SOC Electrodes from Low Resolution Data Using Machine Learning and Computer Vision
Presentation regarding the rapid assessment and optimization of SOC electrodes from low-resolution data using machine learning and computer vision, using training data generated by NETL's Joule supercomputer and high-performance models.
Epting, William↗