SEARCH · Search NASA

Results for “matching problem”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 541 records · Page 30

All sources NASA NTRS DOE OSTI

NASA NTRS1985

Development of an Expert System for Representing Procedural Knowledge

A high level of automation is of paramount importance in most space operations. It is critical for unmanned missions and greatly increases the effectiveness of manned missions. However, although many functions can be automated by using advanced engineering techniques, others require complex reasoning, sensing, and manipulatory capabilities that go beyond this technology. Automation of fault diagnosis and malfunction handling is a case in point. The military have long been interested in this problem, and have developed automatic test equipment to aid in the maintenance of complex military hardware. These systems are all based on conventional software and engineering techniques. However, the effectiveness of such test equipment is severely limited. The equipment is inflexible and unresponsive to the skill level of the technicians using it. The diagnostic procedures cannot be matched to the exigencies of the current situation nor can they cope with reconfiguration or modification of the items under test. The diagnosis cannot be guided by useful advice from technicians and, when a fault cannot be isolated, no explanation is given as to the cause of failure. Because these systems perform a prescribed sequence of tests, they cannot utilize knowledge of a particular situation to focus attention on more likely trouble spots. Consequently, real-time performance is highly unsatisfactory. Furthermore, the cost of developing test software is substantial and time to maturation is excessive. Significant advances in artificial intelligence (AI) have recently led to the development of powerful and flexible reasoning systems, known as expert or knowledge-based systems. We have devised a powerful and theoretically sound scheme for representing and reasoning about procedural knowledge.

Georgeff, Michael P.↗

NASA NTRS1999

Adaptive Error Estimation in Linearized Ocean General Circulation Models

Data assimilation methods are routinely used in oceanography. The statistics of the model and measurement errors need to be specified a priori. This study addresses the problem of estimating model and measurement error statistics from observations. We start by testing innovation based methods of adaptive error estimation with low-dimensional models in the North Pacific (5-60 deg N, 132-252 deg E) to TOPEX/POSEIDON (TIP) sea level anomaly data, acoustic tomography data from the ATOC project, and the MIT General Circulation Model (GCM). A reduced state linear model that describes large scale internal (baroclinic) error dynamics is used. The methods are shown to be sensitive to the initial guess for the error statistics and the type of observations. A new off-line approach is developed, the covariance matching approach (CMA), where covariance matrices of model-data residuals are "matched" to their theoretical expectations using familiar least squares methods. This method uses observations directly instead of the innovations sequence and is shown to be related to the MT method and the method of Fu et al. (1993). Twin experiments using the same linearized MIT GCM suggest that altimetric data are ill-suited to the estimation of internal GCM errors, but that such estimates can in theory be obtained using acoustic data. The CMA is then applied to T/P sea level anomaly data and a linearization of a global GFDL GCM which uses two vertical modes. We show that the CMA method can be used with a global model and a global data set, and that the estimates of the error statistics are robust. We show that the fraction of the GCM-T/P residual variance explained by the model error is larger than that derived in Fukumori et al.(1999) with the method of Fu et al.(1993). Most of the model error is explained by the barotropic mode. However, we find that impact of the change in the error statistics on the data assimilation estimates is very small. This is explained by the large representation error, i.e. the dominance of the mesoscale eddies in the T/P signal, which are not part of the 21 by 1" GCM. Therefore, the impact of the observations on the assimilation is very small even after the adjustment of the error statistics. This work demonstrates that simult&neous estimation of the model and measurement error statistics for data assimilation with global ocean data sets and linearized GCMs is possible. However, the error covariance estimation problem is in general highly underdetermined, much more so than the state estimation problem. In other words there exist a very large number of statistical models that can be made consistent with the available data. Therefore, methods for obtaining quantitative error estimates, powerful though they may be, cannot replace physical insight. Used in the right context, as a tool for guiding the choice of a small number of model error parameters, covariance matching can be a useful addition to the repertory of tools available to oceanographers.

Chechelnitsky, Michael Y.↗

NASA NTRS2016

Development of a Perfectly Matched Layer Technique for a Discontinuous-Galerkin Spectral-Element Method

The numerical simulation of many aerodynamic non-periodic flows of practical interest involves discretized computational domains that often must be artificially truncated. Appropriate boundary conditions are required at these truncated domain boundaries, and ideally, these boundary conditions should be perfectly "absorbing" or "nonreflecting" so that they do not contaminate the flow field in the interior of the domain. The proper specification of these boundaries is critical to the stability, accuracy, convergence, and quality of the numerical solution, and has been the topic of considerable research. The need for accurate boundary specification has been underscored in recent years with efforts to apply higher-fidelity methods (DNS, LES) in conjunction with high-order low-dissipation numerical schemes to realistic flow configurations. One of the most popular choices for specifying these boundaries is the characteristics-based boundary condition where the linearized flow field at the boundaries are decomposed into characteristic waves using either one-dimensional Riemann or other multi-dimensional Riemann approximations. The values of incoming characteristics are then suitably modified. The incoming characteristics are specified at the in flow boundaries, and at the out flow boundaries the variation of the incoming characteristic is zeroed out to ensure no reflection. This, however, makes the problem ill-posed requiring the use of an ad-hoc parameter to allow small reflections that make the solution stable. Generally speaking, such boundary conditions work reasonably well when the characteristic flow direction is normal to the boundary, but reflects spurious energy otherwise. An alternative to the characteristic-based boundary condition is to add additional "buffer" regions to the main computational domain near the artificial boundaries, and solve a different set of equations in the buffer region in order to minimize acoustic reflections. One approach that has been used involves modeling the pressure fluctuations as acoustic waves propagating in the far-field relative to a single noise-source inside the buffer region. This approach treats vorticity-induced pressure fluctuations the same as acoustic waves. Another popular approach, often referred to as the "sponge layer," attempts to dampen the flow perturbations by introducing artificial dissipation in the buffer region. Although the artificial dissipation removes all perturbations inside the sponge layer, incoming waves are still reflected from the interface boundary between the computational domain and the sponge layer. The effect of these refkections can be somewhat mitigated by appropriately selecting the artificial dissipation strength and the extent of the sponge layer. One of the most promising variants on the buffer region approach is the Perfectly Matched Layer (PML) technique. The PML technique mitigates spurious reflections from boundaries and interfaces by dampening the perturbation modes inside the buffer region such that their eigenfunctions remain unchanged. The technique was first developed by Berenger for application to problems involving electromagnetic wave propagation. It was later extended to the linearized Euler, Euler and Navier-Stokes equations by Hu and his coauthors. The PML technique ensures the no-reflection property for all waves, irrespective of incidence angle, wavelength, and propagation direction. Although the technique requires the solution of a set of auxiliary equations, the computational overhead is easily justified since it allows smaller domain sizes and can provide better accuracy, stability, and convergence of the numerical solution. In this paper, the PML technique is developed in the context of a high-order spectral-element Discontinuous Galerkin (DG) method. The technique is compared to other approaches to treating the in flow and out flow boundary, such as those based on using characteristic boundary conditions and sponge layers. The superiority of the current PML technique over other approaches is demonstrated for a range of test cases, viz., acoustic pulse propagation, convective vortex, shear layer flow, and low-pressure turbine cascade flow. The paper is structured as follows. We first derive the PML equations from the non{linear Euler equations. A short description of the higher-order DG method used is then described. Preliminary results for the four test cases considered are then presented and discussed. Details regarding current work that will be included in the final paper are also provided.

Spectral↗

NASA NTRS1983

High frequency green function for aerodynamic noise in moving media. I - General theory. II - Noise from a spreading jet

It is shown how a high frequency analysis can be made for general problems involving flow-generated noise. In the parallel shear flow problem treated by Balsa (1976) and Goldstein (1982), the equation governing sound propagation in the moving medium could be transformed into a wave equation for a stationary medium with an inhomogeneous index of refraction. It is noted that the procedure of Avila and Keller (1963) was then used to construct a high frequency Green function. This procedure involves matching a solution valid in an inner region around the point source to an outer, ray-acoustics solution. This same procedure is used here to construct the Green function for a source in an arbitrary mean flow. In view of the fact that there is no restriction to parallel flow, the governing equations cannot be transformed into a wave equation; the analysis therefore proceeds from the equations of motion themselves.

Durbin, P. A.↗

NASA NTRS2001

A Long-Term Space Astrophysics Research Program: The Evolution of the Quasar Continuum

Four papers have been written. One reports on the major study funded by this grant: a pan-chromatic study of the quasar continuum at redshift 3. Two others make use of the quasar continuum shapes to find the minimum total accretion luminosity of the Universe, and hence the efficiency and spin of supermassive black holes; the second shows that the reemission of absorbed quasar radiation alleviates a major problem with galaxy formation and the FIR background. The last paper recognizes the role quasars may play in the initial formation of dust in the early Universe. The major study of a sample of z=3 and its comparison with a sample of z=0.l quasars across the whole X-ray to radio spectrum was completed and accepted for publication in ApJ Supplements. This study comprises the thesis work of Olga Kuhn. The two samples are matched in evolved luminosity, and so should be sampling the same black hole population at different z, and in different accretion states. Despite this no strong differences were found between the samples, except in the 'small bump' region of the optical/UV. This region is dominated by FeII emission, and may indicate abundance evolution in quasars. The lack of overall spectral changes argues strongly against a single population of quasars fading over cosmic time, and for a multiple generation, or multiple outburst model for quasars. A study of the total luminosity absorbed from quasars and re-emitted in the infrared produced two results (reported in two papers): The minimum intrinsic luminosity/Gpc(3) from AGN compared with the measured mass density in supermassive black holes [Gpc(-3)] requires a conversion efficiency of accreted mass into luminosity of greater than 15%. Non-rotating black holes cannot exceed 5% efficiency, while rapidly rotating black holes can reach 47%. Hence our result requires that most supermassive black holes must be rapidly rotating. The second result comes from considering the contribution that the re-radiated quasar radiation makes to the far infrared background (FIRB). The effective temperature of the re radiation is tightly constrained, but the detailed shape (e.g. line emission, range of temperature) is only of second order importance. At least 15perhaps 20-25% of the FIR background must come from AGN. This contribution significantly relieves problems in galaxy evolution that come from trying to use only starlight to make the FIRB. The third paper addresses the origin of the dust obscuration that is so widespread in AGN. The standard assumption is that the dust comes from the normal star-formation processes in galaxies and is drawn close to the nucleus along with the gas that powers the accretion. In complete contrast, using a wind outflow model for the broad emission line (BEL) region (Elvis 2000) as a basis, we show that BEL clouds will expand, cool and form dust as they flow outward, in strict analogy to the stellar winds of red supergiants.

Elvis, M.↗

NASA NTRS2003

Data-Mining Toolset Developed for Determining Turbine Engine Part Life Consumption

The current practice in aerospace turbine engine maintenance is to remove components defined as life-limited parts after a fixed time, on the basis of a predetermined number of flight cycles. Under this schedule-based maintenance practice, the worst-case usage scenario is used to determine the usable life of the component. As shown, this practice often requires removing a part before its useful life is fully consumed, thus leading to higher maintenance cost. To address this issue, the NASA Glenn Research Center, in a collaborative effort with Pratt & Whitney, has developed a generic modular toolset that uses data-mining technology to parameterize life usage models for maintenance purposes. The toolset enables a "condition-based" maintenance approach, where parts are removed on the basis of the cumulative history of the severity of operation they have experienced. The toolset uses data-mining technology to tune life-consumption models on the basis of operating and maintenance histories. The flight operating conditions, represented by measured variables within the engine, are correlated with repair records for the engines, generating a relationship between the operating condition of the part and its service life. As shown, with the condition-based maintenance approach, the lifelimited part is in service until its usable life is fully consumed. This approach will lower maintenance costs while maintaining the safety of the propulsion system. The toolset is a modular program that is easily customizable by users. First, appropriate parametric damage accumulation models, which will be functions of engine variables, must be defined. The tool then optimizes the models to match the historical data by computing an effective-cycle metric that reduces the unexplained variability in component life due to each damage mode by accounting for the variability in operational severity. The damage increment due to operating conditions experienced during each flight is used to compute the effective cycles and ultimately the replacement time. Utilities to handle data problems, such as gaps in the flight data records, are included in the toolset. The tool was demonstrated using the first stage, high-pressure turbine blade of the PW4077 engine (Pratt & Whitney, East Hartford, CT). The damage modes considered were thermomechanical fatigue and oxidation/erosion. Each PW4077 engine contains 82 first-stage, high-pressure turbine blades, and data from a fleet of engines were used to tune the life-consumption models. The models took into account not only measured variables within the engine, but also unmeasured variables such as engine health parameters that are affected by degradation of the engine due to aging. The tool proved effective at predicting the average number of blades scrapped over time due to each damage mode, per engine, given the operating history of the engine. The customizable tools are available to interested parties within the aerospace community.

Litt, Jonathan S.↗

NASA NTRS2008

Sparse Regression as a Sparse Eigenvalue Problem

We extend the l0-norm "subspectral" algorithms for sparse-LDA [5] and sparse-PCA [6] to general quadratic costs such as MSE in linear (kernel) regression. The resulting "Sparse Least Squares" (SLS) problem is also NP-hard, by way of its equivalence to a rank-1 sparse eigenvalue problem (e.g., binary sparse-LDA [7]). Specifically, for a general quadratic cost we use a highly-efficient technique for direct eigenvalue computation using partitioned matrix inverses which leads to dramatic x103 speed-ups over standard eigenvalue decomposition. This increased efficiency mitigates the O(n4) scaling behaviour that up to now has limited the previous algorithms' utility for high-dimensional learning problems. Moreover, the new computation prioritizes the role of the less-myopic backward elimination stage which becomes more efficient than forward selection. Similarly, branch-and-bound search for Exact Sparse Least Squares (ESLS) also benefits from partitioned matrix inverse techniques. Our Greedy Sparse Least Squares (GSLS) generalizes Natarajan's algorithm [9] also known as Order-Recursive Matching Pursuit (ORMP). Specifically, the forward half of GSLS is exactly equivalent to ORMP but more efficient. By including the backward pass, which only doubles the computation, we can achieve lower MSE than ORMP. Experimental comparisons to the state-of-the-art LARS algorithm [3] show forward-GSLS is faster, more accurate and more flexible in terms of choice of regularization

Exact Sparse Least Squares (ESLS)↗

NASA NTRS1980

Sound propagation through parallel jets exhausting from ducts

The method of matched asymptotic expansions is employed to construct the solution for the propagation of sound through parallel jets which exit from long ducts and are surrounded by a uniform parallel stream. Parts of the duct walls are lined with acoustically absorbent material. The small parameter for the expansion is the ratio of the inner jet thickness to the accoustic wavelength. The problem is further simplified when the condition is imposed that the speed of the outer stream, which accounts for the forward motion speed of the ducts, is much smaller than the speeds of the jets. This condition is valid during landing and takeoff operations. Farfield pressure distributions are obtained for the case in which the inner jet is much faster than the outer jet and the case in which the two jets are the same.

Ting, L.↗

NASA NTRS1991

Lunar crane system

In many lunar construction scenarios, mechanical cranes in some form will be indispensible in moving large masses around with various degrees of fine positioning. While thorough experience exists in the use of terrestrial cranes new thinking is required about the design of cranes to be used in extraterrestrial construction. The primary driving force for this new thinking is the need to automate the crane system so that space cranes can be operated as telerobotic machines with a large number of automatic capabilities. This is true because in extraterrestrial construction human resources will need to be critically rationed. The design problems of mechanisms and control systems for a lunar crane must deal with at least two areas of performance. First, the automated crane must be capable of maneuvering a large mass, so that when the mass arrives at the target position there are only small vibrations. Secondly, any residue vibrations must be automatically damped out and a fine positioning must be achieved. For extraterrestrial use there are additional challenges to a crane design - for example, to design a crane system so that it can be transformed for other construction uses. This initial project in crane design does not address such additional issues, although they may be the subject of future CSC research. To date the Center has designed and analyzed many mechanisms. The fundamental problem of trade-offs between passively stabilizing the load and actively controlling the load by actuators was extensively studied. The capability of 3D dynamics modeling now exists for such studies. A scaled model of a lunar crane was set up and it has been most fruitful in providing basic understanding of lunar cranes. Due to an interesting scaling match-up, this scaled model exhibits the load vibration frequencies one would expect in the real lunar case. Using the analytical results achieved to date, a laboratory crane system is now being developed as a test bed for verifying a wide variety of mechanisms and control designs. Future development will be aimed at making the crane system a telerobotic test bed into which external sensors such as computer vision systems, and other small robotic devices such as CSC lunar rovers, will be integrated.

Mikulas, Martin M., Jr.↗

NASA NTRS2013

A Coupled Approach for Structural Damage Detection with Incomplete Measurements

This historical work couples model order reduction, damage detection, dynamic residual/mode shape expansion, and damage extent estimation to overcome the incomplete measurements problem by using an appropriate undamaged structural model. A contribution of this work is the development of a process to estimate the full dynamic residuals using the columns of a spring connectivity matrix obtained by disassembling the structural stiffness matrix. Another contribution is the extension of an eigenvector filtering procedure to produce full-order mode shapes that more closely match the measured active partition of the mode shapes using a set of modified Ritz vectors. The full dynamic residuals and full mode shapes are used as inputs to the minimum rank perturbation theory to provide an estimate of damage location and extent. The issues associated with this process are also discussed as drivers of near-term development activities to understand and improve this approach.

James, George↗

NASA NTRS2013

Translating MAPGEN to ASPEN for MER

This software translates MAPGEN (Europa and APGEN) domains to ASPEN, and the resulting domain can be used to perform planning for the Mars Exploration Rover (MER). In other words, this is a conversion of two distinct planning languages (both declarative and procedural) to a third (declarative) planning language in order to solve the problem of faithful translation from mixed-domain representations into the ASPEN Modeling Language. The MAPGEN planning system is an example of a hybrid procedural/declarative system where the advantages of each are leveraged to produce an effective planner/scheduler for MER tactical planning. The adaptation of the planning system (ASPEN) was investigated, and, with some translation, much of the procedural knowledge encoding is amenable to declarative knowledge encoding. The approach was to compose translators from the core languages used for adapting MAGPEN, which consists of Europa and APGEN. Europa is a constraint- based planner/scheduler where domains are encoded using a declarative model. APGEN is also constraint-based, in that it tracks constraints on resources and states and other variables. Domains are encoded in both constraints and code snippets that execute according to a forward sweep through the plan. Europa and APGEN communicate to each other using proxy activities in APGEN that represent constraints and/or tokens in Europa. The composition of a translator from Europa to ASPEN was fairly straightforward, as ASPEN is also a declarative planning system, and the specific uses of Europa for the MER domain matched ASPEN s native encoding fairly closely. On the other hand, translating from APGEN to ASPEN was considerably more involved. On the surface, the types of activities and resources one encodes in APGEN appear to match oneto- one to the activities, state variables, and resources in ASPEN. But, when looking into the definitions of how resources are profiled and activities are expanded, one sees code snippets that access various information available during planning for the moment in time being planned to decide at the time what the appropriate profile or expansion is. APGEN is actually a forward (in time) sweeping discrete event simulator, where the model is composed of code snippets that are artfully interleaved by the engine to produce a plan/schedule. To solve this problem, representative code is simulated as a declarative series of task expansions. Predominantly, three types of procedural models were translated: loops, if statements, and code blocks. Loops and if statements were handled using controlled task expansion, and code blocks were handled using constraint networks that maintained the generation of results based on what the order of execution would be for a procedural representation. One advantage with respect to performance for MAPGEN is the use of APGEN s GUI. This GUI is written in C++ and Motif, and performs very well for large plans.

Rabideau, Gregg R.↗

NASA NTRS1998

Low Temperature Photoluminescence Characterization of Orbitally Grown CdZnTe

The II-VI ternary alloy CdZnTe is a technologically important material because of its use as a lattice matched substrate for HgCdTe based devices. The increasingly stringent requirements on performance that must be met by such large area infrared detectors also necessitates a higher quality substrate. Such substrate material is typically grown using the Bridgman technique. Due to the nature of bulk semiconductor growth, gravitationally dependent phenomena can adversely affect crystalline quality. The most direct way to alleviate this problem is by crystal growth in a reduced gravity environment. Since it requires hours, even days, to grow a high quality crystal, an orbiting space shuttle or space station provides a superb platform on which to conduct such research. For well over ten years NASA has been studying the effects of microgravity semiconductor crystal growth. This paper reports the results of photoluminescence characterization performed on an arbitrary grown CdZnTe bulk crystal.

Ritter, Timothy M.↗

NASA NTRS1976

Consideration of design and calibration of terrestrial reference solar cells

A discussion is presented on the problems encountered in the design of a reference cell that meets basic criteria, starting with basic design considerations, and proceeding with the precautions taken to ensure a global monitoring capability. The effects of the variations in atmospheric conditions on the calibration and use of reference cells are presented along with a discussion of the simplifications brought about by the use of spectrally matched test and reference cells. Finally, a method of matching test modules and arrays to reference cells by a red/blue response ratio technique is described.

Weizer, V. G.↗

NASA NTRS1995

Coupling Single-Mode Fiber to Uniform and Symmetrically Tapered Thin-Film Waveguide Structures Using Gadolinium Gallium Garnet

The need for high efficiency components has grown significantly due to the expanding role of fiber optic communications for various applications. Integrated optics is in a state of metamorphosis and there are many problems awaiting solutions. One of the main problems being the lack of a simple and efficient method of coupling single-mode fibers to thin-film devices for integrated optics. In this paper, optical coupling between a single-mode fiber and a uniform and tapered thin-film waveguide is theoretically modeled and analyzed. A novel tapered structure presented in this paper is shown to produce perfect match for power transfer.

Gadi, Jagannath↗

NASA NTRS1997

Determination of Coronal Magnetic Fields from Vector Magnetograms

During the course of the present contract we developed an 'evolutionary technique' for the determination of force-free coronal magnetic fields from vector magnetograph observations. The method can successfully generate nonlinear force- free fields (with non-constant-a) that match vector magnetograms. We demonstrated that it is possible to determine coronal magnetic fields from photospheric measurements, and we applied it to vector magnetograms of active regions. We have also studied theoretical models of coronal fields that lead to disruptions. Specifically, we have demonstrated that the determination of force-free fields from exact boundary data is a well-posed mathematical problem, by verifying that the computed coronal field agrees with an analytic force-free field when boundary data for the analytic field are used; demonstrated that it is possible to determine active-region coronal magnetic fields from photospheric measurements, by computing the coronal field above active region 5747 on 20 October 1989, AR6919 on 15 November 1991, and AR7260 on 18 August 1992, from data taken with the Stokes Polarimeter at Mees Solar Observatory, University of Hawaii; started to analyze active region 7201 on 19 June 1992 using measurements made with the Advanced Stokes Polarimeter at NSO/Sac Peak; investigated the effects of imperfections in the photospheric data on the computed coronal magnetic field; documented the coronal field structure of AR5747 and compared it to the morphology of footpoint emission in a flare, showing that the 'high- pressure' H-alpha footpoints are connected by coronal field lines; shown that the variation of magnetic field strength along current-carrying field lines is significantly different from the variation in a potential field, and that the resulting near-constant area of elementary flux tubes is consistent with observations; begun to develop realistic models of coronal fields which can be used to study flare trigger mechanisms; demonstrated that magnetic nonequilibrium can disrupt sheared coronal arcades, and that helmet streamers can disrupt, leading to coronal mass ejections. Our model has significantly extended the realism with which the coronal magnetic field can be inferred from actual observations. In a subsequent contract awarded by NASA, we have continued to apply and improve the evolutionary technique, to study the physical properties of active regions, and to develop theoretical models of magnetic fields.

Mikic, Zoran↗

NASA NTRS1999

Cultural Variability in Crew Discourse

Four studies were conducted to determine features of effective crew communication in response to errors during flight. Study One examined whether US captains and first officers use different communication strategies to correct errors and problems on the flight deck, and whether their communications are affected by the two situation variables, level of risk and degree of face-threat involved in challenging an error. Study Two was the cross-cultural extension of Study One and involved pilots from three European countries. Study Three compared communication strategies of female and male air carrier pilots who were matched in terms of years and type of aircraft experience. The final study assessed the effectiveness of the communication strategies observed in Study One.

Fischer, Ute↗

NASA NTRS1989

Structural optimization using synthetic functions

An efficient method for solving the structural optimization problem of defining member properties and responses in terms of design variables and calculated direct responses is proposed. The present approach provides the user with a simple means of interfacing design analysis models, as well as imposing constraints on functions which are not otherwise directly available. Results are presented for the examples of a three-bar truss, a 25-bar truss, the matching of calculated responses to measured responses, and the design of a rectangular beam.

Vanderplaats, G. N.↗

NASA NTRS2006

Attenuating Stereo Pixel-Locking via Affine Window Adaptation

For real-time stereo vision systems, the standard method for estimating sub-pixel stereo disparity given an initial integer disparity map involves fitting parabolas to a matching cost function aggregated over rectangular windows. This results in a phenomenon known as 'pixel-locking,' which produces artificially-peaked histograms of sub-pixel disparity. These peaks correspond to the introduction of erroneous ripples or waves in the 3D reconstruction of truly Rat surfaces. Since stereo vision is a common input modality for autonomous vehicles, these inaccuracies can pose a problem for safe, reliable navigation. This paper proposes a new method for sub-pixel stereo disparity estimation, based on ideas from Lucas-Kanade tracking and optical flow, which substantially reduces the pixel-locking effect. In addition, it has the ability to correct much larger initial disparity errors than previous approaches and is more general as it applies not only to the ground plane.

pixel locking effect↗