#### DMCA

## Performance of optical flow techniques (1994)

### Cached

### Download Links

- [www.csee.wvu.edu]
- [www.cs.toronto.edu]
- [www.cs.toronto.edu]
- [www.csd.uwo.ca]
- [orion.scs.carleton.ca]
- [www.csd.uwo.ca]
- [www.cs.duke.edu]
- [www.cs.duke.edu]
- DBLP

### Other Repositories/Bibliography

Venue: | INTERNATIONAL JOURNAL OF COMPUTER VISION |

Citations: | 1315 - 32 self |

### Citations

2847 | An Iterative Image Registration Technique with an Application to Stereo Vision - Lucas, Kanade - 1981 |

2366 | Schunck B.G. Determining optical flow - Horn - 1981 |

1706 | Computer Vision - Horn - 1987 |

1360 | The laplacian pyramid as a compact image code
- Burt, Adelson
- 1983
(Show Context)
Citation Context ... matching strategy. The Laplacian 3 In the original paper [57] the authors used standard deviations of 5 in space and 1 frame in time.sBarron, Fleet and Beauchemin IJCV 12:1, pp43-77, 1994 10 pyramid =-=[13]-=- allows the computation of large displacements between frames and helps to enhance image structure, such as edges, that is often thought to be important. We begin at the coarsest level where displacem... |

1270 | Theory of edge detection
- Marr, Hildreth
- 1980
(Show Context)
Citation Context ...et and Jepson [20, 23]. Waxman, Wu and Bergholm Waxman, Wu and Bergholm [61] apply spatiotemporal lters to binary edge maps to track edges in real-time. Edge maps E(x� t), based on DOG zero-crossings =-=[42]-=-, are smoothed with a Gaussian lter to create a convected activation pro le A(x� t): A(x� t) = G(x�t� x� y� t) E(x� t) : (2.29) Level contours of A(x� t) are then tracked using di erential methods. Ho... |

869 | Spafiotemporal energy models for the perception of motion
- Adelson, Bergen
- 1985
(Show Context)
Citation Context ...echniques is based on the output energy of velocity-tuned lters [2, 7, 11, 27, 30, 34]. These are also called frequency-based methods owing to the design of velocity-tuned lters in the Fourier domain =-=[1, 23, 49, 59]-=-. The Fourier transform of a translating 2-d pattern (2.1) is ^I(k� !) = ^ I 0(k) (! + v T k) � (2.25) where ^ I 0(k) istheFourier transform of I(x� 0), (k) is a Dirac delta function, ! denotes tempor... |

525 |
A Computational Framework and an Algorithm for the Measurement of Visual Motion
- Anandan
- 1989
(Show Context)
Citation Context ...11 In what follows we summarize our main results, concentrating on the techniques that produced reasonably good results, namely, those of Fleet and Jepson [20, 23], Lucas and Kanade [41, 40], Anandan =-=[5, 6]-=-, Uras et al. [57] and Singh [54, 55]. Further quantitative details on the con dence measures can be found in [9]. With respect to rst-order di erential methods, there are several points of interest. ... |

357 | Computation of component image velocity from local phase information - Fleet, Jepson - 1990 |

274 |
An investigation of smoothness constraints for the estimation of displacement vector fields from image sequences
- Nagel, Enkelmann
- 1986
(Show Context)
Citation Context ...cessfully [60, 20]. The second approach uses global smoothness constraints (regularization) in which thevelocity eld is de ned implicitly in terms of the minimum of a functional de ned over the image =-=[32, 43, 44, 46]-=-. Of course, one requirement of di erential techniques is that I(x� t)must be di erentiable. This implies that temporal smoothing at the sensors is needed to avoid aliasing and that numerical di erent... |

233 | Model of human visual-motion sensing
- Watson, Ahumada
- 1985
(Show Context)
Citation Context ...echniques is based on the output energy of velocity-tuned lters [2, 7, 11, 27, 30, 34]. These are also called frequency-based methods owing to the design of velocity-tuned lters in the Fourier domain =-=[1, 23, 49, 59]-=-. The Fourier transform of a translating 2-d pattern (2.1) is ^I(k� !) = ^ I 0(k) (! + v T k) � (2.25) where ^ I 0(k) istheFourier transform of I(x� 0), (k) is a Dirac delta function, ! denotes tempor... |

213 | Probability distributions of optical flow - Simoncelli, Adelson, et al. - 1991 |

204 | Direct passive navigation
- Negahdaripour, Horn
- 1987
(Show Context)
Citation Context ...normal velocity since there is growing interest in the use of normal velocity, thereby side-stepping some of the assumptions inherent incurrent methods for integrating measurements to nd 2-d velocity =-=[3, 4, 10, 16, 33, 47]-=-. We have used both real and synthetic image sequences to test the techniques. In both cases however, we have chosen sequences that are not severely corrupted by spatial or temporal aliasing. This ena... |

172 |
Velocity determination in scenes containing several moving objects
- Fennema, Thompson
- 1979
(Show Context)
Citation Context ...rom spatiotemporal derivatives of image intensity or ltered versions of the image (using low-pass or band-pass lters). The rst instances used rst-order derivatives and were based on image translation =-=[19, 32, 43]-=-, i.e. I(x� t) = I(x ; v t� 0) � (2.1) where v =(u� v) T . From a Taylor expansion of (2.1) [32] or more generally from an assumption that intensity is conserved, dI(x� t)=dt =0,thegradient constraint... |

171 | Optical flow using spatiotemporal filters - Heeger - 1988 |

166 | Model for the extraction of image flow - Heeger - 1987 |

161 | On the estimation of optical flow: Relations between different approaches and some new results - Nagel - 1987 |

144 |
Displacement vectors derived from second-order intensity variations in image sequences
- Nagel
- 1983
(Show Context)
Citation Context ...rom spatiotemporal derivatives of image intensity or ltered versions of the image (using low-pass or band-pass lters). The rst instances used rst-order derivatives and were based on image translation =-=[19, 32, 43]-=-, i.e. I(x� t) = I(x ; v t� 0) � (2.1) where v =(u� v) T . From a Taylor expansion of (2.1) [32] or more generally from an assumption that intensity is conserved, dI(x� t)=dt =0,thegradient constraint... |

134 |
Rotationally invariant image operators
- Beaudet
- 1978
(Show Context)
Citation Context ...nts are then computed by nding the minimum of a quadratic approximation to the SSD surface (about the minimum SSD value found with integer displacements d). As suggested by Anandan, Beaudet operators =-=[12]-=- were used to estimate the quadratic surface parameters. Con dence measures, cmin and cmax, are derived from the principle curvatures, Cmin and Cmax, oftheSSDsurface at the minimum: cmax = Cmax k 1 + ... |

117 |
A computational approach to motion perception
- Uras, Girosi, et al.
- 1988
(Show Context)
Citation Context ...ne techniques including instances of di erential methods, region-based matching, energy-based and phase-based techniques, namely those of Horn and Schunck [32], Lucas and Kanade [40, 41], Uras et al. =-=[57]-=-, Nagel [44], Anandan [5, 6], Singh [54, 55], Heeger [30], Waxman et al. [61] and Fleet and Jepson [20, 23]. Despite their di erences, many of these techniques can be viewed conceptually in terms of t... |

116 |
Measurement of Image Velocity
- Fleet
- 1992
(Show Context)
Citation Context ...-based techniques, namely those of Horn and Schunck [32], Lucas and Kanade [40, 41], Uras et al. [57], Nagel [44], Anandan [5, 6], Singh [54, 55], Heeger [30], Waxman et al. [61] and Fleet and Jepson =-=[20, 23]-=-. Despite their di erences, many of these techniques can be viewed conceptually in terms of three stages of processing: 1. pre ltering or smoothing with low-pass/band-pass lters in order to extract si... |

93 | Stability of phase information - Fleet, Jepson - 1993 |

91 | Optical flow estimation: An error analysis of gradient based methods with local optimazation - Kearney, Thompson, et al. - 1990 |

87 | A Model for the estimate of local image velocity by cells in the visual cortex - Grzywacz, Yuille - 1990 |

81 |
Determining optical ow
- Horn, Schunck
- 1981
(Show Context)
Citation Context ...rage relative errors of 10%{20%, and average angular errors of 7 {12 in the best cases. More recently, Willick andYang [62] have examined the merits of the gradient constraint usedby Horn and Schunck =-=[32]-=- compared to the constraints suggested by Schunck [50, 51] and Nagel [45]. Of these three, they argue that the original gradient constraint is superior. This paper reports a comparison of widely cited... |

78 |
Measuring visual motion from image sequences
- Anandan
- 1987
(Show Context)
Citation Context ...stances of di erential methods, region-based matching, energy-based and phase-based techniques, namely those of Horn and Schunck [32], Lucas and Kanade [40, 41], Uras et al. [57], Nagel [44], Anandan =-=[5, 6]-=-, Singh [54, 55], Heeger [30], Waxman et al. [61] and Fleet and Jepson [20, 23]. Despite their di erences, many of these techniques can be viewed conceptually in terms of three stages of processing: 1... |

77 | Optic flow computation. a unified perspective - Singh - 1992 |

76 |
Surface Orientation and Time to Contact from Image Divergence and Deformation
- Cipolla, Blake
(Show Context)
Citation Context ...normal velocity since there is growing interest in the use of normal velocity, thereby side-stepping some of the assumptions inherent incurrent methods for integrating measurements to nd 2-d velocity =-=[3, 4, 10, 16, 33, 47]-=-. We have used both real and synthetic image sequences to test the techniques. In both cases however, we have chosen sequences that are not severely corrupted by spatial or temporal aliasing. This ena... |

67 |
The extraction of spatio-temporal energy in human and machine vision
- Adelson, Bergen
- 1986
(Show Context)
Citation Context ...ions. Results from both the original and our modi ed method are reported below.sBarron, Fleet and Beauchemin IJCV 12:1, pp43-77, 1994 6 Lucas and Kanade Following Lucas and Kanade [41, 40] and others =-=[2, 37, 52, 53]-=-, we implemented a weighted least-squares (LS) t of local rst-order constraints (2.2) to a constant model for v in each small spatial neighbourhood by minimizing X x2 W 2 (x)[rI(x� t) v + It(x� t)] 2 ... |

67 | Distributed Representation and Analysis of Visual Motion - Simoncelli - 1993 |

64 |
An estimation-theoretic framework for image-flow computation,” in Computer Vision
- Singh
- 1990
(Show Context)
Citation Context ...erential methods, region-based matching, energy-based and phase-based techniques, namely those of Horn and Schunck [32], Lucas and Kanade [40, 41], Uras et al. [57], Nagel [44], Anandan [5, 6], Singh =-=[54, 55]-=-, Heeger [30], Waxman et al. [61] and Fleet and Jepson [20, 23]. Despite their di erences, many of these techniques can be viewed conceptually in terms of three stages of processing: 1. pre ltering or... |

60 | Gneralized image matching by the method of differences - Lucas - 1984 |

58 | Against quantitative optical flow - Verri, Poggio - 1987 |

55 |
Performance of optical ow techniques
- Barron, Fleet, et al.
- 1994
(Show Context)
Citation Context ... 0:21 26.8% Table 4.4: Summary of the Translating Tree 2D Velocity Results. ing the testing of these gradient-based methods and some changes that occurred since we reported our preliminary results in =-=[8, 9]-=-. Our initial implementation quantized the Gaussian smoothed image sequence with 8-bit/pixel for storage, prior to the subsequent gradient computation and least-squares minimization, causing relativel... |

48 |
Velocity estimation from image sequences with second-order dierential operators
- Tretiak, Pastor
- 1984
(Show Context)
Citation Context ...y to solve for both components of v.sBarron, Fleet and Beauchemin IJCV 12:1, pp43-77, 1994 4 Second-order di erential methods use second-order derivatives (the Hessian of I) to constrain 2-d velocity =-=[43, 44, 56, 57]-=-: 2 4 Ixx(x� t) Iyx(x� t) Ixy(x� t) Iyy(x� t) 3 5 0 @ v 1 v 2 1 A + 0 @ Itx(x� t) Ity(x� t) 1 A = 0 @ 0 0 1 A : (2.4) Equation (2.4) can be derived from (2.1) or from the conservation of rI(x� t), drI... |

41 |
On a Constraint Equation for the Estimation of Displacement Rates in Images Sequences
- Nagel
- 1989
(Show Context)
Citation Context ...he best cases. More recently, Willick andYang [62] have examined the merits of the gradient constraint usedby Horn and Schunck [32] compared to the constraints suggested by Schunck [50, 51] and Nagel =-=[45]-=-. Of these three, they argue that the original gradient constraint is superior. This paper reports a comparison of widely cited optical ow methods. We implemented nine techniques including instances o... |

40 |
Multidimensional orientation estimation with applications to texture analysis and optical ow
- Bign, Granlund, et al.
- 1991
(Show Context)
Citation Context ...rom step 1 (2.19), with a threshold based on the largest eigenvalue of Sc (2.20). 2.3 Energy-Based Methods A third class of optical ow techniques is based on the output energy of velocity-tuned lters =-=[2, 7, 11, 27, 30, 34]-=-. These are also called frequency-based methods owing to the design of velocity-tuned lters in the Fourier domain [1, 23, 49, 59]. The Fourier transform of a translating 2-d pattern (2.1) is ^I(k� !) ... |

40 | The computation of the velocity field - Hildreth - 1984 |

34 | Adaptive Multidimensional Filtering
- Haglund
- 1992
(Show Context)
Citation Context ...rom step 1 (2.19), with a threshold based on the largest eigenvalue of Sc (2.20). 2.3 Energy-Based Methods A third class of optical ow techniques is based on the output energy of velocity-tuned lters =-=[2, 7, 11, 27, 30, 34]-=-. These are also called frequency-based methods owing to the design of velocity-tuned lters in the Fourier domain [1, 23, 49, 59]. The Fourier transform of a translating 2-d pattern (2.1) is ^I(k� !) ... |

34 |
Subspace methods for recovering rigid motion
- Heeger, Jepson
- 1992
(Show Context)
Citation Context ...ure require that velocity measurements be accurate and dense, providing a close approximation to the 2-d motion eld. Current techniques require that relative errors in the optical ow be less than 10% =-=[10, 36]-=-. Verri and Poggio [58] have suggested that accurate estimates of the 2-d motion eld are generally inaccessible due to inherent di erences between the 2-d motion eld and intensity variations, while ot... |

33 |
Direct Processing of Curvilinear Sensor Motion from a v(lqtnce of Perspective Images
- J, Brown
- 1984
(Show Context)
Citation Context ...normal velocity since there is growing interest in the use of normal velocity, thereby side-stepping some of the assumptions inherent incurrent methods for integrating measurements to nd 2-d velocity =-=[3, 4, 10, 16, 33, 47]-=-. We have used both real and synthetic image sequences to test the techniques. In both cases however, we have chosen sequences that are not severely corrupted by spatial or temporal aliasing. This ena... |

32 | Multi-resolution flow-through motion analysis - Burt, Yen, et al. - 1983 |

23 |
and A.Waxman. Contour Evolution, Neighbourhood Deformation and Local Image Flow: Curved Surfacesin MOtion
- Wohn
(Show Context)
Citation Context ... accomplish this: The rst method ts the measurements in each neighbourhood to a local model for 2-d velocity (e.g. a low-order polynomial model), using least-squares minimization or a Hough transform =-=[19, 37, 41, 54, 60]-=-. Usually v(x) is taken to be constant, although linear models for v(x) have been used successfully [60, 20]. The second approach uses global smoothness constraints (regularization) in which theveloci... |

22 | Analysis of differential and matching methods for optical flow - Little, Verri - 1988 |

21 | The motion constraints equation for optical flow - Schunck - 1984 |

20 | F.: Convected activation profiles and receptive fields for real time measurement of short range visual motion - Waxman, Wu, et al. - 1988 |

19 |
Active egomotion estimation: a qualitative approach
- Aloimonos, Duric
- 1992
(Show Context)
Citation Context ... Poggio [58] have suggested that accurate estimates of the 2-d motion eld are generally inaccessible due to inherent di erences between the 2-d motion eld and intensity variations, while others (e.g. =-=[4]-=-) argue that the measurement of optical ow is an ill-posed problem. For these reasons it has been suggested that only qualitative information can be extracted. Many methods for computing optical ow ha... |

19 |
A data set for quantitative motion anal- ysis
- Dutta, Manmatha, et al.
- 1989
(Show Context)
Citation Context ...of di erential methods and matching methods on the same data without the complexities of hierarchical coarsene control and warping techniques. For example, we do not consider stop-and-shoot sequences =-=[18]-=-.sBarron, Fleet and Beauchemin IJCV 12:1, pp43-77, 1994 3 This paper concentrates on the accuracy and density ofvelocity estimates produced by the nine methods. Con dence measures have been used to ex... |

18 | Computation of optic flow from the motion of edge features in image sequences - Buxton, Buxton - 1984 |

16 |
Optical ow using spatiotemporal lters
- Heeger
- 1988
(Show Context)
Citation Context ... region-based matching, energy-based and phase-based techniques, namely those of Horn and Schunck [32], Lucas and Kanade [40, 41], Uras et al. [57], Nagel [44], Anandan [5, 6], Singh [54, 55], Heeger =-=[30]-=-, Waxman et al. [61] and Fleet and Jepson [20, 23]. Despite their di erences, many of these techniques can be viewed conceptually in terms of three stages of processing: 1. pre ltering or smoothing wi... |

16 | Parallel optical flow using local voting - Little, Butlhoff, et al. - 1988 |

15 |
Estimation of velocity, acceleration and disparity in time sequences
- Bårman, Haglund, et al.
- 1991
(Show Context)
Citation Context ...rom step 1 (2.19), with a threshold based on the largest eigenvalue of Sc (2.20). 2.3 Energy-Based Methods A third class of optical ow techniques is based on the output energy of velocity-tuned lters =-=[2, 7, 11, 27, 30, 34]-=-. These are also called frequency-based methods owing to the design of velocity-tuned lters in the Fourier domain [1, 23, 49, 59]. The Fourier transform of a translating 2-d pattern (2.1) is ^I(k� !) ... |

15 |
Estimation of depth from motion using anthropomorphic visual sensor
- Tistarelli, Sandini
- 1990
(Show Context)
Citation Context ...ible motion elds. To measure image velocity, assuming drI(x� t)=dt = 0, the constraints in (2.4) may be used in isolation or together with (2.2) to yield an over-determined system of linear equations =-=[24, 48]-=-. However, if the aperture problem prevails in a local neighbourhood (i.e. if intensity is e ectively one-dimensional), then because of the sensitivity ofnumerical di erentiation, 2 nd -order derivati... |

12 | On the estimation of optical ow: Relations between di erent approaches and som new results - Nagel - 1987 |

12 |
Probability distributions of optical ow
- Simoncelli, Adelson, et al.
- 1991
(Show Context)
Citation Context ...ions. Results from both the original and our modi ed method are reported below.sBarron, Fleet and Beauchemin IJCV 12:1, pp43-77, 1994 6 Lucas and Kanade Following Lucas and Kanade [41, 40] and others =-=[2, 37, 52, 53]-=-, we implemented a weighted least-squares (LS) t of local rst-order constraints (2.2) to a constant model for v in each small spatial neighbourhood by minimizing X x2 W 2 (x)[rI(x� t) v + It(x� t)] 2 ... |

11 |
Experimental evaluation of motion constraint equations
- Willick, Yang
- 1991
(Show Context)
Citation Context ...e, synthetic test cases� the accuracy they reported was not encouraging, with average relative errors of 10%{20%, and average angular errors of 7 {12 in the best cases. More recently, Willick andYang =-=[62]-=- have examined the merits of the gradient constraint usedby Horn and Schunck [32] compared to the constraints suggested by Schunck [50, 51] and Nagel [45]. Of these three, they argue that the original... |

11 | Constraints for the computation of optical flow - Girosi, Verri, et al. - 1989 |

9 |
Model for the extraction of image ow
- Heeger
- 1987
(Show Context)
Citation Context ...th our image sequences are close to those for our implementation of the Lucas and Kanade gradient-based method and therefore support this claim. Heeger Here we consider the method developed by Heeger =-=[29, 30]-=-, formulated as a least-squares t of spatiotemporal energy to a plane in frequency space. Local energy is extracted using Gabor-energy lters, with 12 lters at each of several spatial scales, tuned to ... |

9 | Phase singularities in scale-space
- Jepson, Fleet
- 1991
(Show Context)
Citation Context ...ghbourhoods about phase singularities. Such instabilities can be detected with a straightforward constraint on the instantaneous frequency of the lter output and its amplitude variation in space-time =-=[21, 23, 35]-=-: k rlog R(x� t) ; i(k� !) k k � (2.36) where (k�!) denotes the spatiotemporal frequency to which the lter is tuned, k is the standard deviation of the isotropic amplitude spectra they use and denotes... |

8 |
Optical ow estimation: An error analysis of gradient-based methods with local optimization
- Kearney, Thompson, et al.
- 1987
(Show Context)
Citation Context ...o appear. Lacking, however, are quantitative evaluations of existing methods and direct 1sBarron, Fleet and Beauchemin IJCV 12:1, pp43-77, 1994 2 comparisons on a single set of inputs. Kearney et al. =-=[37]-=- discussed sources of error with gradient-based methods. Little and Verri [39] compared properties of di erential and matching methods and reported some quantitative comparisons, but only on two relat... |

8 |
Elaborated Reichardt detectors
- G
- 1985
(Show Context)
Citation Context ...echniques is based on the output energy of velocity-tuned lters [2, 7, 11, 27, 30, 34]. These are also called frequency-based methods owing to the design of velocity-tuned lters in the Fourier domain =-=[1, 23, 49, 59]-=-. The Fourier transform of a translating 2-d pattern (2.1) is ^I(k� !) = ^ I 0(k) (! + v T k) � (2.25) where ^ I 0(k) istheFourier transform of I(x� 0), (k) is a Dirac delta function, ! denotes tempor... |

7 |
Scene matching through hierarchical correlation
- Glazer, Reynolds, et al.
- 1983
(Show Context)
Citation Context ...e a small number of frames exist or because of aliasing in the image acquisition process. In these cases di erential approaches may be inappropriate and it is natural to turn to region-based matching =-=[25, 6, 14, 38, 39]-=-. Such approaches de ne velocity v as the shift d =(dx�dy) that yields the best t between image regions at di erent times. Finding the best match amounts to maximizing a similarity measure (over d), s... |

7 | Against quantitative optical ow - Verri, Poggio - 1987 |

6 |
Image sequence analysis of complex physical objects: nonlinear small scale water surface waves
- Jahne
- 1987
(Show Context)
Citation Context |

6 |
Generalized Image Matching by the Method of Di®erences
- Lucas
- 1984
(Show Context)
Citation Context ...ods. We implemented nine techniques including instances of di erential methods, region-based matching, energy-based and phase-based techniques, namely those of Horn and Schunck [32], Lucas and Kanade =-=[40, 41]-=-, Uras et al. [57], Nagel [44], Anandan [5, 6], Singh [54, 55], Heeger [30], Waxman et al. [61] and Fleet and Jepson [20, 23]. Despite their di erences, many of these techniques can be viewed conceptu... |

5 |
Temporal edges: the detection of motion and the computation of optical
- Duncan, Chou
- 1988
(Show Context)
Citation Context ...We refer to our fourth class of methods as phase-based, because velocity is de ned in terms of the phase behaviour of band-pass lter outputs. For this report we have classi ed zerocrossing techniques =-=[15, 17, 28, 61]-=- as phase-based methods because zero-crossings can be viewed as level phase-crossings. The generalized use of phase information for optical ow was rst developed by Fleet and Jepson [20, 23]. Waxman, W... |

5 |
Analysis of dierential and matching methods for optical
- Little, Verri
- 1989
(Show Context)
Citation Context ...nd direct 1sBarron, Fleet and Beauchemin IJCV 12:1, pp43-77, 1994 2 comparisons on a single set of inputs. Kearney et al. [37] discussed sources of error with gradient-based methods. Little and Verri =-=[39]-=- compared properties of di erential and matching methods and reported some quantitative comparisons, but only on two relatively simple, synthetic test cases� the accuracy they reported was not encoura... |

4 | The feasibility of motion and structure from noisy time-varying image velocity information - D, Tsotsos - 1990 |

4 | Image flow continuity equations for motion and density - Schunck - 1986 |

3 | Direct methods for recovering motion - J - 1988 |

3 |
Parallel optical ow using local voting
- Little, Bultho, et al.
- 1988
(Show Context)
Citation Context ...e a small number of frames exist or because of aliasing in the image acquisition process. In these cases di erential approaches may be inappropriate and it is natural to turn to region-based matching =-=[25, 6, 14, 38, 39]-=-. Such approaches de ne velocity v as the shift d =(dx�dy) that yields the best t between image regions at di erent times. Finding the best match amounts to maximizing a similarity measure (over d), s... |

3 |
The motion constraint equation for optical ow
- Schunck
- 1984
(Show Context)
Citation Context ...rrors of 7 {12 in the best cases. More recently, Willick andYang [62] have examined the merits of the gradient constraint usedby Horn and Schunck [32] compared to the constraints suggested by Schunck =-=[50, 51]-=- and Nagel [45]. Of these three, they argue that the original gradient constraint is superior. This paper reports a comparison of widely cited optical ow methods. We implemented nine techniques includ... |

3 | Optic Flow Computation : A Uni ed Perspective - Singh - 1991 |

3 | Toward real-time optical flow - Fleet, Langley - 1993 |

2 |
Multiresolution ow-through motion analysis
- Burt, Yen, et al.
- 1983
(Show Context)
Citation Context ...e a small number of frames exist or because of aliasing in the image acquisition process. In these cases di erential approaches may be inappropriate and it is natural to turn to region-based matching =-=[25, 6, 14, 38, 39]-=-. Such approaches de ne velocity v as the shift d =(dx�dy) that yields the best t between image regions at di erent times. Finding the best match amounts to maximizing a similarity measure (over d), s... |

2 |
Computation of optic ow from the motion of edge features in image sequences
- Buxton, Buxton
- 1984
(Show Context)
Citation Context ...We refer to our fourth class of methods as phase-based, because velocity is de ned in terms of the phase behaviour of band-pass lter outputs. For this report we have classi ed zerocrossing techniques =-=[15, 17, 28, 61]-=- as phase-based methods because zero-crossings can be viewed as level phase-crossings. The generalized use of phase information for optical ow was rst developed by Fleet and Jepson [20, 23]. Waxman, W... |

2 | Constraints for the computation of optical ow - Girosi, Verri, et al. - 1989 |

2 |
Verri A
- Little
- 1989
(Show Context)
Citation Context ...ontinue to appear. Lacking however, is quantitative evaluation of existin methods, and direct comparisons. Kearney et al. [22$ discussed sources of error with gradient-based methods. Little and Verri =-=[23]-=- compared properties of differential and matching methods, and reported some quantitative comparisons, but only on two relatively simple, synthetic test cases; the accuracy they reported was not encou... |

1 |
Computation of component imagevelocity from local phase information
- Fleet, Jepson
- 1990
(Show Context)
Citation Context ...-based techniques, namely those of Horn and Schunck [32], Lucas and Kanade [40, 41], Uras et al. [57], Nagel [44], Anandan [5, 6], Singh [54, 55], Heeger [30], Waxman et al. [61] and Fleet and Jepson =-=[20, 23]-=-. Despite their di erences, many of these techniques can be viewed conceptually in terms of three stages of processing: 1. pre ltering or smoothing with low-pass/band-pass lters in order to extract si... |

1 |
Toward real-time optical ow
- Fleet, Langley
- 1993
(Show Context)
Citation Context ... 0:375� 0:25� 0:0625) as in [52]. The temporal support forsBarron, Fleet and Beauchemin IJCV 12:1, pp43-77, 1994 7 the entire process was 15 frames. In a more recent implementation, Fleet and Langley =-=[22]-=- have replaced the FIR lters with IIR recursive lters and temporally recursive estimation. This method requires only three frames of storage, delays of only 2 or 3 frames, and yields results of simila... |

1 |
The computation of the velocity eld.Proc
- Hildreth
- 1984
(Show Context)
Citation Context ...We refer to our fourth class of methods as phase-based, because velocity is de ned in terms of the phase behaviour of band-pass lter outputs. For this report we have classi ed zerocrossing techniques =-=[15, 17, 28, 61]-=- as phase-based methods because zero-crossings can be viewed as level phase-crossings. The generalized use of phase information for optical ow was rst developed by Fleet and Jepson [20, 23]. Waxman, W... |

1 |
Image ow continuity equations for motion and density
- Schunck
- 1986
(Show Context)
Citation Context ...rrors of 7 {12 in the best cases. More recently, Willick andYang [62] have examined the merits of the gradient constraint usedby Horn and Schunck [32] compared to the constraints suggested by Schunck =-=[50, 51]-=- and Nagel [45]. Of these three, they argue that the original gradient constraint is superior. This paper reports a comparison of widely cited optical ow methods. We implemented nine techniques includ... |

1 |
Convected activation pro les and receptive elds for real time measurement of short range visual motion
- Waxman, Wu, et al.
- 1988
(Show Context)
Citation Context ...ng, energy-based and phase-based techniques, namely those of Horn and Schunck [32], Lucas and Kanade [40, 41], Uras et al. [57], Nagel [44], Anandan [5, 6], Singh [54, 55], Heeger [30], Waxman et al. =-=[61]-=- and Fleet and Jepson [20, 23]. Despite their di erences, many of these techniques can be viewed conceptually in terms of three stages of processing: 1. pre ltering or smoothing with low-pass/band-pas... |

1 | JOSA A2 - Adelson, Bergen - 1985 |

1 | Adelson E - Burt - 1983 |

1 | Measurement ofImage Velocity - Fleet - 1992 |

1 | Heeger D - Jepson - 1990 |

1 | Boley D - Kearney, Thompson - 1987 |

1 | and Sperling G.: JOSA - van |

1 |
Heeger D
- Simoncelli, Adelson
- 1991
(Show Context)
Citation Context ...tial neighbourhoods R were 5 x 5 pixels, and the window function W(x) was separable and isotropic; its 1-d weights in the horizontal and vertical directions were (0.0625,0.25,0.375,0.25,0.0625) as in =-=[28]-=-. The temporal support for the entire process is 15 frames. Simoncelli et al. [28] present a Bayesian perspective of (5). .Their MAP solution is similar to (6), and yields confidence measures for the ... |

1 |
Poggio T
- Verri
- 1987
(Show Context)
Citation Context ...ion of egomotion and surface structure require accurate and dense velocity measurements; current techniques require that relative errors in the optical flow be less than 10% [6, 211. Verri and Poggio =-=[32]-=- suggest that accurate estimates of the motion field are generally inaccessible. Many methods for computing optical flow have been proposed - others continue to appear. Lacking however, is quantitativ... |

1 | Authorized licensed use limited to - Waxman, Wu, et al. - 1988 |