17.4 Error Analysis, Evaluation for Optical Flow

Barron, J.L., Fleet, D.J., and Beauchemin, S.S.,
Performance of Optical Flow Techniques,
IJCV(12), No. 1, February 1994, pp. 43-77.
Springer DOI Link
HTML Version.
WWW Version. BibRef 9402
And: Add: Burkitt, T.A., CVPR92(236-242).
IEEE Abstract. IEEE Top Reference. Code, Optic Flow. Survey, Optic Flow. Survey of the field and a comparison of a variety of techniques. Compares quality of results, not execution time. Compares: Lucas/Kanade ( See also Generalized Image Matching by the Method of Differences. ), Fleet/Jepson ( See also Hierarchial Construction of Orientation and Velocity Selective Filters. ), Uras ( See also Computational Approach to Motion Perception, A. ), Nagel ( See also On a Constraint Equation for the Estimation of Displacement Rates in Image Sequences. ), Anandan ( See also Computational Framework and an Algorithm for the Measurement of Visual Motion, A. ), Horn/Shunck ( See also Determining Optical Flow. ), Singh ( See also Image-Flow Computation: An Estimation-Theoretic Framework and a Unified Perspective. ). Code for all of these is available from:
WWW Version. BibRef

Kearney, J.K., Thompson, W.B., and Boley, D.L.,
Optical Flow Estimation: An Error Analysis of Gradient-Based Methods with Local Optimization,
PAMI(9), No. 2, March 1987, pp. 229-244. BibRef 8703
Earlier:
Gradient Based Estimation of Disparity,
PRIP82(246-251). Optical Flow, Evaluation. Develops a framework to evaluate errors in local gradient based techniques. Analyze the sources of errors to see which can be addressed. Errors include highly textured regions, non locally constant image flows, and poor variation in the brightness gradient. There is a good summary of optical flow techniques in the paper. BibRef

Kearney, J.K.,
Gradient-Based Estimation of Optical-Flow,
Ph.D.Thesis (CS), Univ. of MN, 1983. BibRef 8300

Kearney, J.K., Thompson, W.B.,
Gradient-Based Estimation of Optical Flow with Global Optimization,
CAIA84(376-380). BibRef 8400

Adiv, G.,
Inherent Ambiguities in Recovering 3-D Motion and Structure from a Noisy Flow Field,
PAMI(11), No. 5, May 1989, pp. 477-489.
IEEE Abstract. IEEE Top Reference.
WWW Version. BibRef 8905
Earlier: CVPR85(70-77). BibRef
And: DARPA85(399-412) in a longer version. (UMass) Noise causes errors (ambiguity) which is inherent, but some parameters are still valid. Nothing surprising in the general conclusions. See also Determining 3-D Motion and Structure from Optical Flow Generated by Several Moving Objects. BibRef

Jasinschi, R.S.,
Intrinsic Constraints in Space-Time Filtering: A New Approach to Representing Uncertainity in Low-Level Vision,
PAMI(14), No. 3, March 1992, pp. 353-366.
IEEE Abstract. IEEE Top Reference.
WWW Version. BibRef 9203

Jasinschi, R.S.,
The Properties of Space-Time Sampling and the Extraction of the Optical Flow: The Effects of Motion Uncertainty,
JVCIR(2), 1991, pp. 222-229. BibRef 9100
Earlier:
Space-Time Sampling with Motion Uncertainty: Constraints on Space-Time Filtering,
ICCV88(428-434).
IEEE Abstract. IEEE Top Reference. BibRef

Jasinschi, R.[Rado], Yuille, A.[Alan],
Non-Rigid Motion and Regge Calculus,
MIT AI Memo-996, November 1987. BibRef 8711

Holt, R.J., Netravali, A.N.,
Motion from Optic Flow: Multiplicity of Solutions,
JVCIR(4), 1993, pp. 14-24. See also Number of Solutions for Motion and Structure from Multiple Frame Correspondence. BibRef 9300

Ben-Tzvi, D., del Bimbo, A., Nesi, P.,
Optical Flow from Constraint Lines Parametrization,
PR(26), No. 10, October 1993, pp. 1549-1561.
WWW Version. BibRef 9310

Nesi, P., del Bimbo, A., Ben-Tzvi, D.,
A Robust Algorithm for Optical-Flow Estimation,
CVIU(62), No. 1, July 1995, pp. 59-68.
WWW Version. See also Algorithms for Optical Flow Estimation in Real-Time on Connection Machine-2. BibRef 9507

del Bimbo, A., Nesi, P., Sanz, J.L.C.,
Optical flow computation using extended constraints,
IP(5), No. 5, May 1996, pp. 720-739.
IEEE DOI Link 0402
BibRef

del Bimbo, A., Nesi, P., and Sanz, J.L.C.,
Analysis of Optical-Flow Constraints,
IP(4), No. 4, April 1995, pp. 460-469.
IEEE DOI Link BibRef 9504
Earlier: Univ. of FlorenceTR. Analysis of constraints. BibRef

Denney, Jr., T.S., Prince, J.L.,
A frequency domain performance analysis of Horn and Schunck's optical flow algorithm for deformable motion,
IP(4), No. 9, September 1995, pp. 1324-1327.
IEEE DOI Link 0402
See also Determining Optical Flow. BibRef

Denney, Jr., T.S.,
On estimating 3-D incompressible motion,
ICIP95(III: 492-495).
IEEE DOI Link 9510
BibRef

Gupta, S.N., Prince, J.L.,
Stochastic formulations of optical flow algorithms under variable brightness conditions,
ICIP95(III: 484-487).
IEEE DOI Link 9510
BibRef

Earnshaw, A.M.[A. Mark], Blostein, S.D.[Steven D.],
The Performance of Camera Translation Direction Estimators from Optical-flow: Analysis, Comparison, and Theoretical Limits,
PAMI(18), No. 9, September 1996, pp. 927-932.
IEEE Abstract. IEEE Top Reference.
WWW Version. Translation Estimation. Linear Constraints. BibRef 9609
Earlier:
An error analysis of camera translation direction estimation from optical flow using linear constraints,
ICIP95(I: 394-397).
IEEE DOI Link 9510
BibRef
And:
The Performance of Camera Translation Direction Estimators,
TR- 95-10, Dept. ECE, Queens Univ. 1995. BibRef

Yang, J., Stevenson, S.B.,
Effects of Spatial-Frequency, Duration, and Contrast on Discriminating Motion Directions,
JOSA-A(14), No. 9, September 1997, pp. 2041-2048. 9709
BibRef

Fermüller, C.[Cornelia], Aloimonos, Y.[Yannis],
Ambiguity in Structure from Motion: Sphere Versus Plane,
IJCV(28), No. 2, June/July 1998, pp. 137-154.
WWW Version. 9808
BibRef
Earlier:
What Is Computed by Structure from Motion Algorithms?,
ECCV98(I: 359).
WWW Version. BibRef
And: UMD--TR3809, June 1997. Structure from Motion. Normal Flow. Error Analysis.
WWW Version.
WWW Version. BibRef

Fermüller, C.[Cornelia], Aloimonos, Y.[Yiannis],
Observability of 3D Motion,
IJCV(37), No. 1, June 2000, pp. 43-63.
WWW Version. 0005
BibRef
Earlier:
Which Shape from Motion?,
ICCV98(689-695).
IEEE DOI Link BibRef

Fermüller, C., Aloimonos, Y.,
The Confounding of Translation and Rotation in Reconstruction from Multiple Views,
CVPR97(250-256).
IEEE Abstract. IEEE Top Reference.
WWW Version. 9704
Given optic flow, what are optimal relations of T and R? BibRef

Fermüller, C.[Cornelia], Aloimonos, Y.[Yiannis],
Analysis of Reconstruction from Multiple Views,
DARPA97(1411-1418). BibRef 9700

Fermüller, C.[Cornelia], and Aloimonos, Y.[Yiannis],
Algorithm-Independent Stability Analysis of Structure from Motion,
UMDTR3691, September 1996.
WWW Version.
WWW Version. Analysis of flow along some direction and spherical retina. BibRef 9609

Liu, H.C.[Hong-Che], Hong, T.H.[Tsai-Hong], Herman, M., Camus, T.A., Chellappa, R.,
Accuracy vs. Efficiency Trade-Offs in Optical Flow Algorithms,
CVIU(72), No. 3, December 1998, pp. 271-286.
WWW Version. BibRef 9812
Earlier: A1, A2, A3, A5 Only: ECCV96(II:174-183).
Springer DOI Link Comparison of different techniques. Liu ( See also Generalized Motion Model for Estimating Optical Flow Using 3-D Hermite Polynomials, A. ), Lucas/Kanade ( See also Generalized Image Matching by the Method of Differences. ), Fleet/Jepson ( See also Hierarchial Construction of Orientation and Velocity Selective Filters. ), Anandan ( See also Computational Framework and an Algorithm for the Measurement of Visual Motion, A. ), Camus ( See also Real-Time Quantized Optical Flow. ), Nagel ( See also On a Constraint Equation for the Estimation of Displacement Rates in Image Sequences. ), Uras ( See also Computational Approach to Motion Perception, A. ), Horn/Shunck ( See also Determining Optical Flow. ). BibRef

Fleury, M., Clark, A.F., Downton, A.C.,
Evaluating optical-flow algorithms on a parallel machine,
IVC(19), No. 3, February 2001, pp. 131-143.
WWW Version. 0103
BibRef

Fermüller, C.[Cornelia], Shulman, D.[David], Aloimonos, Y.[Yiannis],
The Statistics of Optical Flow,
CVIU(82), No. 1, April 2001, pp. 1-32.
WWW Version. 0001
BibRef
And: But: A3 is: Pless, R.[Robert], UMD--TR4080, November 1999.
WWW Version.
WWW Version. See also On the Geometry of Visual Correspondence. BibRef

Fermüller, C., Aloimonos, Y.,
The Statistics of Optical Flow: Implications for the Process of Correspondence in Vision,
ICPR00(Vol I: 119-126).
IEEE DOI Link
HTML Version. 0009
BibRef

Fermüller, C.[Cornelia], Pless, R.[Robert], Aloimonos, Y.[Yiannis],
Statistical Biases in Optic Flow,
CVPR99(I: 561-566).
IEEE Abstract. IEEE Top Reference.
WWW Version. BibRef 9900

Fermüller, C.[Cornelia], Baker, P.[Patrick], Aloimonos, Y.[Yiannis],
Visual space-time geometry: A tool for perception and the imagination,
PIEEE(90), No. 7, July 2002, pp. 1113-1135.
IEEE DOI Link 0207
BibRef
Earlier:
Geometry and Statistics of Visual Space-Time,
VI02(53).
PDF Version. 0208
BibRef

Dror, R.O.[Ron O.], O'Carroll, D.C.[David C.], Laughlin, S.B.[Simon B.],
Accuracy of velocity estimation by Reichardt correlators,
JOSA-A(18), No. 2, February 2001, pp. 241-252. 0102
BibRef

Mizukami, Y.[Yoshiki], Sato, T.[Taiji], Tanaka, K.[Kanya],
A Comparison Study for Displacement Computation: Horn and Shunck's method versus March's method,
PRL(22), No. 6-7, May 2001, pp. 825-831.
HTML Version. 0105
See also Determining Optical Flow. See also Computation of Stereo Disparity Using Regularization. BibRef

Ng, L.[Lydia], Solo, V.[Victor],
Errors-in-variables modeling in optical flow estimation,
IP(10), No. 10, October 2001, pp. 1528-1540.
IEEE DOI Link 0110
BibRef
Earlier:
Optical Flow Estimation using Adaptive Wavelet Zeroing,
ICIP99(III:722-726).
IEEE Abstract. IEEE Top Reference. BibRef
Earlier:
Choosing the optimal neighbourhood size in optical flow problems with errors-in-variables modelling,
ICIP98(II: 186-190).
IEEE DOI Link 9810
BibRef
Earlier:
A Data-Driven Method for Choosing Smoothing Parameters in Optical Flow Problems,
ICIP97(III: 360-363).
IEEE DOI Link BibRef

Solo, V.,
A sure-fired way to choose smoothing parameters in ill-conditioned inverse problems,
ICIP96(III: 89-92).
IEEE DOI Link 9610
BibRef

Dougherty, L., Asmuth, J.C., Blom, A.S., Axel, L., Kumar, R.,
Validation of an optical flow method for tag displacement estimation,
MedImg(18), No. 4, April 1999, pp. 359-363.
IEEE Top Reference. 0110
BibRef

McCane, B.[Brendan], Novins, K.[Kevin], Crannitch, D.[Dion], and Galvin, B.[Ben],
On Benchmarking Optical Flow,
CVIU(84), No. 1, October 2001, pp. 126-143.
WWW Version. 0203
BibRef

McCane, B.[Brendan],
Optic Flow Evaluation,
OnlineMarch 2007.
WWW Version. Code, Optic Flow. BibRef 0703

Galvin, B., McCane, B., Novins, K., Mason, D., Mills, S.,
Recovering Motion Fields: An Evaluation of Eight Optical Flow Algorithms,
BMVC98(xx-yy). BibRef 9800

Langer, M.S.[Michael S.], Mann, R.[Richard],
Optical Snow,
IJCV(55), No. 1, September 2003, pp. 55-71.
WWW Version. 0307
BibRef
Earlier:
Tracking through Optical Snow,
BMCV02(181 ff.).
HTML Version. 0303
BibRef
Earlier:
Dimensional Analysis of Image Motion,
ICCV01(I: 155-162).
IEEE DOI Link 0106
Optical snow -- motion through a very cluttered scene. See also Spectrum analysis of motion parallax in a 3D cluttered scene and application to egomotion. BibRef

Chapdelaine-Couture, V., Roy, S., Langer, M.S., Mann, R.,
Principal Components Analysis of Optical Snow,
BMVC04(xx-yy).
HTML Version. 0508
BibRef

Mann, R.[Richard], Langer, M.S.[Michael S.],
Optical flow and the aperture problem,
ICPR02(IV: 264-267).
IEEE DOI Link 0211
BibRef

Lim, S.H.[Suk-Hwan], Apostolopoulos, J.G., Gamal, A.E.,
Optical flow estimation using temporally oversampled video,
IP(14), No. 8, August 2005, pp. 1074-1087.
IEEE DOI Link 0508
BibRef
Earlier:
Benefits of temporal oversampling in optical flow estimation,
ICIP04(IV: 2567-2570).
IEEE DOI Link 0505
BibRef

Ohta, N.[Naoya], Nishizawa, S.[Satoe],
How Much Does Color Information Help Optical Flow Computation?,
IEICE(E89-D), No. 5, May 2006, pp. 1759-1762.
WWW Version. 0605
BibRef

Sun, D.Q.[De-Qing], Roth, S.[Stefan], Lewis, J.P., Black, M.J.[Michael J.],
Learning Optical Flow,
ECCV08(III: 83-97).
Springer DOI Link 0810
BibRef

Baker, S.[Simon], Roth, S.[Stefan], Scharstein, D.[Daniel], Black, M.J.[Michael J.], Lewis, J.P., Szeliski, R.[Richard],
A Database and Evaluation Methodology for Optical Flow,
ICCV07(1-8).
IEEE DOI Link 0710
Dataset, Optical Flow. BibRef

Ulman, V.[Vladimír], Hubený, J.[Jan],
Pseudo-real Image Sequence Generator for Optical Flow Computations,
SCIA07(976-985).
Springer DOI Link 0706
Generation of artificial flow data for evaluation of computations. BibRef

Adachi, E.[Eisuke], Kurita, T.[Takio], Otsu, N.[Nobuyuki],
Reliability index of optical flow that considers error margin of matches and stabilizes camera movement estimation,
ICPR06(I: 699-702).
WWW Version. 0609
BibRef

Austvoll, I.[Ivar],
A Study of the Yosemite Sequence Used as a Test Sequence for Estimation of Optical Flow,
SCIA05(659-668).
Springer DOI Link 0506
BibRef

Spies, H., Barron, J.L.,
Evaluating certainties in image intensity differentation for optical flow,
CRV04(408-416).
IEEE Abstract. IEEE Top Reference. 0408
BibRef
Earlier:
Evaluating the range flow motion constraint,
ICPR02(III: 517-520).
IEEE DOI Link 0211
BibRef

Okada, R., Maki, A., Taniguchi, Y., Onoguchi, K.,
Temporally evaluated optical flow: study on accuracy,
ICPR02(I: 343-347).
IEEE DOI Link 0211
BibRef

Mester, R.,
A system-theoretical view on local motion estimation,
Southwest02(201-205).
IEEE Top Reference. 0208
BibRef

Zhao, W., Sawhney, H.S.,
Is Super-Resolution with Optical Flow Feasible?,
ECCV02(I: 599 ff.).
HTML Version. 0205
BibRef

Olson, C.F.[Clark F.], Matthies, L.H.[Larry H.], Schoppers, M.J.[Marcel J.], Maimone, M.W.[Mark W.],
Robust Stereo Ego-Motion for Long Distance Navigation,
CVPR00(II: 453-458).
IEEE Abstract. IEEE Top Reference.
WWW Version. 0005
Add orientation sensor; then position error is reduced BibRef

Nishimura, T., Oka, R., Held, A., Kojima, H.,
Effect of Time-Spatial Size of Motion Image for Localization by Using the Spotting Method,
ICPR96(I: 191-195).
IEEE DOI Link 9608
(Information Integration Lab. J) BibRef

Shaw, G.B.,
Determining Motion Parameters Using a Perturbation Approach,
COINSTR 83-30, UMass., September 1983. now at Univ of Oregon, Discusses the error problems of traditional methods for computation of motion parameters from noisy data, and proposes a solution. BibRef 8309

Little, J.J., and Verri, A.,
Analysis of Differential and Matching Methods for Optical Flow,
Motion89(173-180). BibRef 8900
And: MIT AI Memo-1066, August 1988. Optical Flow, Evaluation. BibRef

Denzler, J.[Joachim], Schless, V., Paulus, D., Niemann, H.,
Statistical Approach to Classification of Flow Patterns for Motion Detection,
ICIP96(I: 517-520).
IEEE DOI Link
Postscript Version. BibRef 9600

Niemann, H., Arnold, J., Sagerer, G.,
On the Accuracy of Optical Flow Computation Using Global Optimization,
ICPR88(II: 1094-1096).
IEEE DOI Link
IEEE Top Reference. BibRef 8800

Chapter on Optical Flow Field Computations and Use continues in
Optical Flow Field -- Smoothness .