## Symmetrical Dense Optical Flow Estimation with Occlusion Detection (2002)

Venue: | In ECCV |

Citations: | 46 - 1 self |

### BibTeX

@INPROCEEDINGS{Alvarez02symmetricaldense,

author = {Luis Alvarez and Rachid Deriche and Théo Papadopoulo and Javier Sanchez},

title = {Symmetrical Dense Optical Flow Estimation with Occlusion Detection},

booktitle = {In ECCV},

year = {2002},

pages = {721--735},

publisher = {Springer}

}

### Years of Citing Articles

### OpenURL

### Abstract

Traditional techniques of dense optical flowestimation don't generally yield symmetrical solutions: the results will di#er if they are applied between images I1 and I2 or between images I2 and I1 . In this work, we present a method to recover a dense optical flow field map from two images, while explicitely taking into account the symmetry across the images as well as possible occlusions and discontinuities in the flow field. The idea is to consider both displacements vectors from I1 to I2 and I2 to I1 and to minimise an energy functional that explicitely encodes all those properties. This variational problem is then solved using the gradient flowdefined by theEuler--Lw7458) equations associated to the energy. In order to reduce the risk to be trapped within some irrelevant minimum, a focusing strategy based on a multi-resolution technique is used to converge toward the solution. Promising experimental results on both synthetic and real images are presented to illustrate the capabilities of this symmetrical variational approach to recover accurate optical flow. 1

### Citations

1819 | Determining optical flow
- Horn, Schunck
- 1981
(Show Context)
Citation Context ... are usuallyrequired. The most used one is to add a quadratic constraint on the gradient magnitude of the flow to impose some form of smoothness on the flow field as done originallyby Horn and Schunk =-=[16]. -=-Other formulations have been proposed using different smoothness constraints, but despite their interest, theyclearlylacked robustness A. Heyden et al. (Eds.): ECCV 2002, LNCS 2350, pp. 721–735, 200... |

1089 | Performance of optical flow techniques
- Barron, Fleet, et al.
(Show Context)
Citation Context ...n extensivelyaddressed in the literature, and the most common methods are correlation, gradient, spatiotemporal filtering, Fourier phase and energy based approaches. We refer the interested reader to =-=[6,17]-=- for an excellent analysis and evaluation of important optical flow methods. Manyof these approaches use the classical constraint equation that relates the gradient of brightness to the components u a... |

327 |
Image selective smoothing and edge detection by nonlinear diusion (II
- Alvarez, Lions, et al.
- 1992
(Show Context)
Citation Context ...sotropicallyalong the edge. This behaviour is verysimilar to edge-enhancing anisotropic diffusion filtering [29,28], and it is also close in spirit to the modified mean-curvature motion considered in =-=[2,1]. At t-=-his point, two remarks can be made: – Since Ω1 is usuallyunknown, the domain D1 is used instead in the previous integrals. – The criterion depicted above is clearlynon-symmetrical in the images ... |

251 |
Geometric Partial Differential Equations and Image Analysis
- Sapiro
- 2001
(Show Context)
Citation Context ...nt minima [20,4,3]. The method presented here is inspired from this kind of variational framework that has also proven recentlyto be veryuseful in manyother image processing and computer vision tasks =-=[25,30,8,1]-=-. We start explicitelywith the method described in [4] and modifyit using ideas similar to those developed in [9] to obtain an algorithm that will give the same solution if applied between images A an... |

236 |
Anisotropic Diffusion
- Weickert
- 1998
(Show Context)
Citation Context ...nt minima [20,4,3]. The method presented here is inspired from this kind of variational framework that has also proven recentlyto be veryuseful in manyother image processing and computer vision tasks =-=[25,30,8,1]-=-. We start explicitelywith the method described in [4] and modifyit using ideas similar to those developed in [9] to obtain an algorithm that will give the same solution if applied between images A an... |

226 |
An investigation of smoothness constraints for the estimation of displacement vector fields from image sequences
- Nagel, Enkelmann
- 1986
(Show Context)
Citation Context ...r to estimate the optical flow more accurately, one have to explicitely take into account the problem of occlusions and discontinuities. See the following works, mostlybased on variational approaches =-=[20,19,10,24,15,14,22,13,5,18,11,3,31]-=-. Due to the fact that the functional to be minimised is generallynot convex, some focusing strategyembedding the method in a multi-resolution scheme or a linear scale-space have been successfullyappl... |

208 | Anisotropic Diffusion in Image Processing
- Weickert
- 1998
(Show Context)
Citation Context ...we have isotropic behaviour within regions, and at image boundaries the process smoothes anisotropicallyalong the edge. This behaviour is verysimilar to edge-enhancing anisotropic diffusion filtering =-=[29,28], an-=-d it is also close in spirit to the modified mean-curvature motion considered in [2,1]. At this point, two remarks can be made: – Since Ω1 is usuallyunknown, the domain D1 is used instead in the p... |

192 | On the unification of line processes, outlier rejection, and robust statistics with applications to early vision - Black, Rangarajan - 1996 |

137 |
On the Estimation of Optical Flow: Relations Between Different Approaches and Some New Results
- Nagel
- 1987
(Show Context)
Citation Context ...e algorithm across images boundaries, we have chosen to use Nagel-Enkelmann operator [20] , first because its efficiencyhas been demonstrated numerous times in the context of optical flow estimations =-=[3,6,12, 19,21,20,26,27] and because of its simplici-=-tysince the underlying second order differential operator is linear. E R � 1 (h1) = with: Ω1 Φ (D(�∇I1�), ∇h1) dm1 , Φ (D(∇Ii), ∇hi) =trace � ∇(hi) t D(∇Ii)∇(hi) � , where ... |

124 |
Multimodal estimation of discontinuous optical flow using Markov random fields
- Heitz, Bouthemy
- 1993
(Show Context)
Citation Context ...r to estimate the optical flow more accurately, one have to explicitely take into account the problem of occlusions and discontinuities. See the following works, mostlybased on variational approaches =-=[20,19,10,24,15,14,22,13,5,18,11,3,31]-=-. Due to the fact that the functional to be minimised is generallynot convex, some focusing strategyembedding the method in a multi-resolution scheme or a linear scale-space have been successfullyappl... |

101 | Reliable estimation of dense optical flow fields with large displacements
- Alvarez, Weickert, et al.
(Show Context)
Citation Context ...lynot convex, some focusing strategyembedding the method in a multi-resolution scheme or a linear scale-space have been successfullyapplied to reduce the risk to get trapped in some irrelevant minima =-=[20,4,3]-=-. The method presented here is inspired from this kind of variational framework that has also proven recentlyto be veryuseful in manyother image processing and computer vision tasks [25,30,8,1]. We st... |

93 |
Investigations of multigrid algorithms for the estima- tion of optical flow fields in image sequences
- Enkelmann
- 1988
(Show Context)
Citation Context ...e algorithm across images boundaries, we have chosen to use Nagel-Enkelmann operator [20] , first because its efficiencyhas been demonstrated numerous times in the context of optical flow estimations =-=[3,6,12, 19,21,20,26,27] and because of its simplici-=-tysince the underlying second order differential operator is linear. E R � 1 (h1) = with: Ω1 Φ (D(�∇I1�), ∇h1) dm1 , Φ (D(∇Ii), ∇hi) =trace � ∇(hi) t D(∇Ii)∇(hi) � , where ... |

79 | A theoretical framework for convex regularizers in PDE-based computation of image motion,” Univ
- Weickert, Schnörr
- 2000
(Show Context)
Citation Context ...r to estimate the optical flow more accurately, one have to explicitely take into account the problem of occlusions and discontinuities. See the following works, mostlybased on variational approaches =-=[20,19,10,24,15,14,22,13,5,18,11,3,31]-=-. Due to the fact that the functional to be minimised is generallynot convex, some focusing strategyembedding the method in a multi-resolution scheme or a linear scale-space have been successfullyappl... |

64 | Dense disparity map estimation respecting image discontinuities: A pde and scale-space based approach
- Alvarez, Deriche, et al.
- 2002
(Show Context)
Citation Context ...nt minima [20,4,3]. The method presented here is inspired from this kind of variational framework that has also proven recentlyto be veryuseful in manyother image processing and computer vision tasks =-=[25,30,8,1]-=-. We start explicitelywith the method described in [4] and modifyit using ideas similar to those developed in [9] to obtain an algorithm that will give the same solution if applied between images A an... |

54 | Consistent linear elastic transformations for image matching
- Christensen
- 1999
(Show Context)
Citation Context ...centlyto be veryuseful in manyother image processing and computer vision tasks [25,30,8,1]. We start explicitelywith the method described in [4] and modifyit using ideas similar to those developed in =-=[9]-=- to obtain an algorithm that will give the same solution if applied between images A and B or between images B and A. Contrarilyto [9], we deal with discontinuities of the flow field and explicitelyta... |

52 |
Optical flow estimation: advances and comparisons
- Otte, Nagel
- 1994
(Show Context)
Citation Context ...◦ h2 − Id� (left) and �h2 ◦ h1 − Id� (right). 4.2 Real Images The algorithm depicted above has also been run on various pairs of real images. One particularlyuseful pair is that provided=-= in the paper [23]-=- as it provides the ground truth optical flow. These images have been kindlyprovided bythe KOGS/IAKS group of Karlsruhe University. Figures 7 and 8 show the input images and the two components of the ... |

49 | On benchmarking optical flow
- McCane, Novins, et al.
- 2001
(Show Context)
Citation Context ...n extensivelyaddressed in the literature, and the most common methods are correlation, gradient, spatiotemporal filtering, Fourier phase and energy based approaches. We refer the interested reader to =-=[6,17]-=- for an excellent analysis and evaluation of important optical flow methods. Manyof these approaches use the classical constraint equation that relates the gradient of brightness to the components u a... |

48 | Computing optical flow via variational techniques
- Aubert, Deriche, et al.
- 1999
(Show Context)
Citation Context |

46 | Theoretical foundations of anisotropic diffusion in image processing
- Weickert
- 1996
(Show Context)
Citation Context ...we have isotropic behaviour within regions, and at image boundaries the process smoothes anisotropicallyalong the edge. This behaviour is verysimilar to edge-enhancing anisotropic diffusion filtering =-=[29,28], an-=-d it is also close in spirit to the modified mean-curvature motion considered in [2,1]. At this point, two remarks can be made: – Since Ω1 is usuallyunknown, the domain D1 is used instead in the p... |

40 | Constraints for the estimation of displacement vector fields from image sequences
- Nagel
- 1983
(Show Context)
Citation Context |

37 | A scale-space approach to nonlocal optical flow calculations
- Alvarez, Weickert, et al.
- 1999
(Show Context)
Citation Context |

32 |
Introduction to the special issue on partial differential equations and geometry-driven diffusion in image processing and analysis
- Caselles, Morel, et al.
- 2009
(Show Context)
Citation Context |

31 | Nonlinear variational method for optical flow computation
- Cohen
- 1993
(Show Context)
Citation Context |

31 |
A fast scalable algorithm for discontinuous optical flow estimation
- Ghosal, Vanek
- 1996
(Show Context)
Citation Context |

23 |
Variational approach to optical flow estimation managing discontinuities
- Nesi
- 1993
(Show Context)
Citation Context |

21 |
Optical flow estimation while preserving its discontinuities: A variational approach
- Deriche, Kornprobst, et al.
- 1995
(Show Context)
Citation Context |

2 |
Accurate estimation of discontinuous optical flow by minimizing divergence related functionals
- Guichard, Rudin
- 1996
(Show Context)
Citation Context |

2 |
On the mathematical fundations of smoothness constraints for the determination of optical flow and for surface reconstruction
- Snyder
- 1995
(Show Context)
Citation Context ...e algorithm across images boundaries, we have chosen to use Nagel-Enkelmann operator [20] , first because its efficiencyhas been demonstrated numerous times in the context of optical flow estimations =-=[3,6,12, 19,21,20,26,27] and because of its simplici-=-tysince the underlying second order differential operator is linear. E R � 1 (h1) = with: Ω1 Φ (D(�∇I1�), ∇h1) dm1 , Φ (D(∇Ii), ∇hi) =trace � ∇(hi) t D(∇Ii)∇(hi) � , where ... |

1 |
Multiresolution Markov random field and multigrid algorithm for discontinuity preserving estimation of the optical flow
- Memin, Perez
- 1995
(Show Context)
Citation Context |

1 |
Determination of Optical Flowand its Discontinuities using Non-Linear Diffusion
- Proesmans, Gool, et al.
- 1994
(Show Context)
Citation Context |

1 |
Determining optical flowfor irregular domains by minimizing quadratic functionals of a certain class
- Schnörr
- 1991
(Show Context)
Citation Context |