Search NASA⌕ Search

SEARCH · Search NASA

Results for “Corner detection”

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 19 records

Rapid Corner Detection Using FPGAs

In order to perform precision landings for space missions, a control system must be accurate to within ten meters. Feature detection applied against images taken during descent and correlated against the provided base image is computationally expensive and requires tens of seconds of processing time to do just one image while the goal is to process multiple images per second. To solve this problem, this algorithm takes that processing load from the central processing unit (CPU) and gives it to a reconfigurable field programmable gate array (FPGA), which is able to compute data in parallel at very high clock speeds. The workload of the processor then becomes simpler; to read an image from a camera, it is transferred into the FPGA, and the results are read back from the FPGA. The Harris Corner Detector uses the determinant and trace to find a corner score, with each step of the computation occurring on independent clock cycles. Essentially, the image is converted into an x and y derivative map. Once three lines of pixel information have been queued up, valid pixel derivatives are clocked into the product and averaging phase of the pipeline. Each x and y derivative is squared against itself, as well as the product of the ix and iy derivative, and each value is stored in a WxN size buffer, where W represents the size of the integration window and N is the width of the image. In this particular case, a window size of 5 was chosen, and the image is 640 480. Over a WxN size window, an equidistance Gaussian is applied (to bring out the stronger corners), and then each value in the entire window is summed and stored. The required components of the equation are in place, and it is just a matter of taking the determinant and trace. It should be noted that the trace is being weighted by a constant k, a value that is found empirically to be within 0.04 to 0.15 (and in this implementation is 0.05). The constant k determines the number of corners available to be compared against a threshold sigma to mark a valid corner. After a fixed delay from when the first pixel is clocked in (to fill the pipeline), a score is achieved after each successive clock. This score corresponds with an (x,y) location within the image. If the score is higher than the predetermined threshold sigma, then a flag is set high and the location is recorded.

Morfopoulos, Arin C.↗

Dye Penetrant Crack Detectability in External Corners in Presence of Fillet Radius

NASA uses special dye penetrant nondestructive evaluation process to provide reliable detection of very small cracks. Typically the surface crack lengths sizes are 0.030' and 0.050' for special dye penetrant process. Qualification requires demonstration of crack detection on a set of cracks with average crack size smaller than or equal to the qualification crack size. The demonstration is called point estimate demonstration. A set of corner cracks can be used to determine reliably detectable corner crack using the point estimate demonstration method. However, dye penetrant demonstration on surface cracks can be used to assess reliably detectable corner crack sizes by using similarity in the penetrant process. The paper provides similarity analysis approach for determining the reliably detectable corner crack sizes for given a point estimate demonstrated surface crack size.

Koshti, Ajay M.↗

Adaptive Filtering and Edge Detection Using Range Data

It is typical in edge and corner detection applications to examine a single scale or to consider some space of scales in the image without knowing which scale is appropriate for each location in the image.

Image smoothing edge detection stereoscopy filteri↗

Developing Methods for Exercise System Kinematic Tracking

BACKGROUND How to quantify the load and forces produced by exercise equipment and their Vibration Isolation and Stabilization (VIS) platforms in-flight is an active area of investigation. Kinematic tracking paired with system modeling can provide insights as well as verification and validation of simulations used for system design and development. Traditional motion capture methods can require significant cost in equipment procurement and crew-time, but newer lessons learned can be leveraged [1]. The VIS systems of current and future exercise hardware on the International Space Station (ISS) such as the Cycle Ergometer with Vibration Isolation System (CEVIS) and the European Enhanced Exploration Exercise Device (E4D) are not currently outfitted with IMUs or similar measurement devices. Video-based methods would enable use of multi-purpose, crew-familiar flight equipment. An initial exploration of video-based solutions was performed utilizing 2-camera video from crew cycling on Teal-CEVIS on the ISS. METHODS AND RESULTS Our group has scoped a variety of video-based object tracking methods. To date, we have primarily investigated computer vision toolkits such as open CV. Techniques explored include key-point detection, background subtraction, region-of interest tracking, color-based tracking, tag masking and tracking, and corner detection. Although object-tracking and 6D pose estimation is a rich field, space applications are a unique problem that are challenging for existing software and toolkits. The majority of the existing object-tracking applications involve vehicles/pedestrians and household objects with simple backgrounds. We have identified the following features which pose particular challenges for on-station exercise equipment tracking: 1. Busy and visually cluttered background 2. Low-textured tracking object with relatively small motions 3. Occlusions and motion by human subject and loose, floating objects 4. Limited number of video cameras with no fixed global references 5. Limited ability to add tags, markers, or visual references to the tracking object 6. Lack of training data for Machine Learning (ML) algorithms CONCLUSION We will summarize the efficacy of techniques tested for a ground mock-trial and the on-station exercise trial. It is likely that human-in-loop feedback or a conglomerate of methods is required. ML-based methods, like those implemented for human body tracking [2], may still be a viable option, but more training data and validation is needed.

L Nilsson↗

Developing Methods for Exercise System Kinematics Tracking

BACKGROUND How to quantify the load and forces produced by exercise equipment and their Vibration Isolation and Stabilization (VIS) platforms in-flight is an active area of investigation. Kinematic tracking paired with system modeling can provide insights as well as verification and validation of simulations used for system design and development. Traditional motion capture methods can require significant cost in equipment procurement and crew-time, but newer lessons learned can be leveraged [1]. The VIS systems of current and future exercise hardware on the International Space Station (ISS) such as the Cycle Ergometer with Vibration Isolation System (CEVIS) and the European Enhanced Exploration Exercise Device (E4D) are not currently outfitted with IMUs or similar measurement devices. Video-based methods would enable use of multi-purpose, crew-familiar flight equipment. An initial exploration of video-based solutions was performed utilizing 2-camera video from crew cycling on Teal-CEVIS on the ISS. METHODS AND RESULTS Our group has scoped a variety of video-based object tracking methods. To date, we have primarily investigated computer vision toolkits such as openCV. Techniques explored include key-point detection, background subtraction, region-of interest tracking, color-based tracking, tag masking and tracking, and corner detection. Although object-tracking and 6D pose estimation is a rich field, space applications are a unique problem that are challenging for existing software and toolkits. The majority of the existing object-tracking applications involve vehicles/pedestrians and household objects with simple backgrounds. We have identified the following features which pose particular challenges for on-station exercise equipment tracking: Busy and visually cluttered background Low-textured tracking object with relatively small motions Occlusions and motion by human subject and loose, floating objects Limited number of video cameras with no fixed global references Limited ability to add tags, markers, or visual references to the tracking object Lack of training data for Machine Learning (ML) algorithms CONCLUSION We will summarize the efficacy of techniques tested for a ground mock-trial and the on-station exercise trial. It is likely that human-in-loop feedback or a conglomerate of methods is required. ML-based methods, like those implemented for human body tracking [2], may still be a viable option, but more training data and validation is needed.

L B Nilsson↗

Image processing in remote sensing data analysis - The state of the art

Image analysis techniques applicable to remote sensing data and covering image models, feature detection, segmentation and classification, texture analysis, and matching are studied. Model types for characterizing images examined include random-field, mosaic, and facet models. Edge and corner detection as well as global extraction of linear features are discussed. Pixel clustering and classification are covered in addition to the regional approach to segmentation. Autocorrelation, second-order gray level probability density, and the use of primitive element statistics are discussed in relation to texture analysis. Finally, reducing the cost of (sub)imaging matching methods (e.g., pixelwise comparison of gray levels and normalized cross-correlation between two images) as well as improving match sharpness is considered.

Rosenfeld, A.↗

An automated computerized vision technique for determination of three-dimensional object geometry

It is very important to determine three dimensional geometry of objects quickly in various military, space, construction, and industrial applications. An automatic scheme to obtain three dimensional geometry of objects by employing only one camera is presented. At present, this technique is applicable to a limited category of objects, satisfying the following constraints: they are flat-surfaces, and all the vertex points have to be recognized as corner points of the two dimensional image. The scheme consists of corner detection, data communication, camera calibration techniques and point searching and matching, edge cancelation, and creation procedures. An L shaped model is chosen as a test object. Experimental results demonstrated the reconstruction of this object geometry within 5 mm discrepancy. This scheme is quite convenient, efficient to use and can be applied to a wide range of problems in the real world.

Chiang, Pen-Tai↗

NASA Tech Briefs, December 2010

Topics include: Coherent Frequency Reference System for the NASA Deep Space Network; Diamond Heat-Spreader for Submillimeter-Wave Frequency Multipliers; 180-GHz I-Q Second Harmonic Resistive Mixer MMIC; Ultra-Low-Noise W-Band MMIC Detector Modules; 338-GHz Semiconductor Amplifier Module; Power Amplifier Module with 734-mW Continuous Wave Output Power; Multiple Differential-Amplifier MMICs Embedded in Waveguides; Rapid Corner Detection Using FPGAs; Special Component Designs for Differential-Amplifier MMICs; Multi-Stage System for Automatic Target Recognition; Single-Receiver GPS Phase Bias Resolution; Ultra-Wideband Angle-of-Arrival Tracking Systems; Update on Waveguide-Embedded Differential MMIC Amplifiers; Automation Framework for Flight Dynamics Products Generation; Product Operations Status Summary Metrics; Mars Terrain Generation; Application-Controlled Parallel Asynchronous Input/Output Utility; Planetary Image Geometry Library; Propulsion Design With Freeform Fabrication (PDFF); Economical Fabrication of Thick-Section Ceramic Matrix Composites; Process for Making a Noble Metal on Tin Oxide Catalyst; Stacked Corrugated Horn Rings; Refinements in an Mg/MgH2/H2O-Based Hydrogen Generator; Continuous/Batch Mg/MgH2/H2O-Based Hydrogen Generator; Strain System for the Motion Base Shuttle Mission Simulator; Ko Displacement Theory for Structural Shape Predictions; Pyrotechnic Actuator for Retracting Tubes Between MSL Subsystems; Surface-Enhanced X-Ray Fluorescence; Infrared Sensor on Unmanned Aircraft Transmits Time-Critical Wildfire Data; and Slopes To Prevent Trapping of Bubbles in Microfluidic Channels.

Source record↗

Finding corner point correspondence from wavelet decomposition of image data

A time efficient algorithm for image registration between two images that differ in translation is discussed. The algorithm is based on coarse-fine strategy using wavelet decomposition of both the images. The wavelet decomposition serves two different purposes: (1) its high frequency components are used to detect feature points (corner points here) and (2) it provides coarse-to-fine structure for making the algorithm time efficient. The algorithm is based on detecting the corner points from one of the images called reference image and computing corresponding points from the other image called test image by using local correlations using 7x7 windows centered around the corner points. The corresponding points are detected at the lowest decomposition level in a search area of about 11x11 (depending on the translation) and potential points of correspondence are projected onto higher levels. In the subsequent levels the local correlations are computed in a search area of no more than 3x3 for refinement of the correspondence.

Manohar, Mareboyana↗

Remote measurements of ambient air pollutants with a bistatic laser system

The ambient air pollutants ozone, nitric oxide, and ethylene have been monitored in the Pasadena area with a bistatic IR laser apparatus. These pollutants were measured with a differential absorption technique, using selected wavelengths in the 9.5-, 5.2-, and 10.5-micron regions, respectively. The transmitted laser radiation was detected using both direct and heterodyne detection techniques. In the direct detection case, cube corner retroreflectors provided the return, and the heterodyne detection responded to scattered radiation from various rough surfaces, ranging from 400 m to 1.9 km in distance from the apparatus. Significant departures from ambient background concentration levels were noticed in the region near a local freeway during periods of moderate and heavy traffic.

Menzies, R. T.↗

Guidelines for Mapping Reliably Detectable Dye Penetrant Crack Size at External Corners with Fillet Radii

The technical memorandum addresses some concerns with dye penetrant crack sizes stated in Table 1 (and Table 2) of NASA-STD-5009B. Table 1 corner crack size of a = 0.100” and c = 0.150” is reasoned to be larger than it should be based on the original Standard dye penetrant qualification. In the original Standard dye penetrant qualification, surface crack size with depth a = 0.075” x length 2c = 0.150” was qualified with minimum 90% Probability of Detection (POD) with 95% confidence (conf.). No corner crack size was qualified by direct POD testing in original dye penetrant qualification. Standard dye penetrant detectable crack size for radiused corner is not addressed in NASA-STD-5009B.

Dye Penetrant Testing↗

An investigation of the flow characteristics in the blade endwall corner region

Studies were undertaken to determine the structure of the flow in the blade end wall corner region simulated by attaching two uncambered airfoils on either side of a flat plate with a semicircular leading edge. Detailed measurements of the corner flow were obtained with conventional pressure probes, hot wire anemometry, and flow visualization. The mean velocity profiles and six components of the Reynolds stress tensor were obtained with an inclined single sensor hot wire probe whereas power spectra were obtained with a single sensor oriented normal to the flow. Three streamwise vortices were identified based on the surface streamlines, distortion of total pressure profiles, and variation of mean velocity components in the corner. A horseshoe vortex formed near the leading edge of the airfoil. Within a short distance downstream, a corner vortex was detected between the horseshoe vortex and the surfaces forming the corner. A third vortex was formed at the rear portion of the corner between the corner vortex and the surface of the flat plate. Turbulent shear stress and production of turbulence are negligibly small. A region of negative turbulent shear stress was also observed near the region of low turbulence intensity from the vicinity of the flat plate.

Hazarika, Birinchi K.↗

Detecting Negative Obstacles by Use of Radar

Robotic land vehicles would be equipped with small radar systems to detect negative obstacles, according to a proposal. The term "negative obstacles" denotes holes, ditches, and any other terrain features characterized by abrupt steep downslopes that could be hazardous for vehicles. Video cameras and other optically based obstacle-avoidance sensors now installed on some robotic vehicles cannot detect obstacles under adverse lighting conditions. Even under favorable lighting conditions, they cannot detect negative obstacles. A radar system according to the proposal would be of the frequency-modulation/ continuous-wave (FM/CW) type. It would be installed on a vehicle, facing forward, possibly with a downward slant of the main lobe(s) of the radar beam(s) (see figure). It would utilize one or more wavelength(s) of the order of centimeters. Because such wavelengths are comparable to the characteristic dimensions of terrain features associated with negative hazards, a significant amount of diffraction would occur at such features. In effect, the diffraction would afford a limited ability to see corners and to see around corners. Hence, the system might utilize diffraction to detect corners associated with negative obstacles. At the time of reporting the information for this article, preliminary analyses of diffraction at simple negative obstacles had been performed, but an explicit description of how the system would utilize diffraction was not available.

Mittskus, Anthony↗

Development and flight test of a weather radar precision approach concept

In order to make full use of the helicopter's unique capability of remote-site, off-airport landings, it would be desirable to employ a self-contained navigation system requiring minimum groundable-based equipment. For this reason, research is being conducted with the aim to develop the use of airborne weather radar as a primary navigation aid for helicopter approach and landing in instrument flight rules (IFR) conditions. Anderson et al. (1982) have reported about the first phase of this effort, taking into account the detection of passive ground-based corner reflectors with the aid of an 'echo processor'. The technology of passive-reflector detection in the overland environment provides the pilot with the range and bearing to the landing site. The present investigation is concerned with a second research phase, which was undertaken with the objective to develop and demonstrate the feasibility of a weather radar-based precision approach concept. Preliminary flight test results are considered.

Clary, G. R.↗

Cracks in Flow Liners and Their Resolution

Cracks were detected in flow liners at the gimbal joints in the LH2 feedlines of the space shuttle's main engines. The cracks initiated at defects in the drainage slots of the flow liners and grew due to high cycle fatigue. Fracture mechanics analyses were conducted to evaluate the life of the liners. These analyses yielded extremely short lives in the presence of small surface or corner cracks. A high fidelity detection method, edge replication, was used to detect the very small cracks. The detected cracks were removed by polishing and the surface quality of the slots was reestablished to improve life of the liners.

Harris, C. E.↗

Research on techniques for laser ranging to optical corner reflectors on the moon. Research on laser techniques and single photo-electron detection and timing

Experimental studies using a pulsed LED, Cerenkov source, and a 100 ps laser were made of various photomultipliers and discriminator combinations. In addition, a new type of neodymium-YAG frequency doubled laser was used as the basis for the development of a stable, short pulse, high repetition rate laser system. This laser was then used in conjunction with atomic clocks to study the effect of gravitational potential on elapsed time. Avenues to promote the development of international cooperation in the area of lunar laser ranging were also explored.

Alley, C. O.↗

Robot acting on moving bodies (RAMBO): Preliminary results

A robot system called RAMBO is being developed. It is equipped with a camera, which, given a sequence of simple tasks, can perform these tasks on a moving object. RAMBO is given a complete geometric model of the object. A low level vision module extracts and groups characteristic features in images of the object. The positions of the object are determined in a sequence of images, and a motion estimate of the object is obtained. This motion estimate is used to plan trajectories of the robot tool to relative locations nearby the object sufficient for achieving the tasks. More specifically, low level vision uses parallel algorithms for image enchancement by symmetric nearest neighbor filtering, edge detection by local gradient operators, and corner extraction by sector filtering. The object pose estimation is a Hough transform method accumulating position hypotheses obtained by matching triples of image features (corners) to triples of model features. To maximize computing speed, the estimate of the position in space of a triple of features is obtained by decomposing its perspective view into a product of rotations and a scaled orthographic projection. This allows the use of 2-D lookup tables at each stage of the decomposition. The position hypotheses for each possible match of model feature triples and image feature triples are calculated in parallel. Trajectory planning combines heuristic and dynamic programming techniques. Then trajectories are created using parametric cubic splines between initial and goal trajectories. All the parallel algorithms run on a Connection Machine CM-2 with 16K processors.

Davis, Larry S.↗