## 3-D Motion and Structure from 2-D Motion Causally Integrated over Time: Implementation (2000)

Venue: | In IEEE Trans. Robotics and Automation |

Citations: | 20 - 1 self |

### BibTeX

@INPROCEEDINGS{Chiuso003-dmotion,

author = {Alessandro Chiuso and Paolo Favaro and Hailin Jin and Stefano Soatto},

title = {3-D Motion and Structure from 2-D Motion Causally Integrated over Time: Implementation},

booktitle = {In IEEE Trans. Robotics and Automation},

year = {2000},

pages = {734--750}

}

### Years of Citing Articles

### OpenURL

### Abstract

The causal estimation of three-dimensional motion from a sequence of two-dimensional images can be posed as a nonlinear filtering problem. We describe the implementation of an algorithm whose uniform observability, minimal realization and stability have been proven analytically in [5]. We discuss a scheme for handling occlusions, drift in the scale factor and tuning of the lter. We also present an extension to partially calibrated camera models and prove its observability. We report the performance of our implementation on a few long sequences of real images. More importantly, however, we have made our real-time implementation -- which runs on a personal computer -- available to the public for first-hand testing.

### Citations

1985 | T.: An iterative image registration technique with an application to stereo vision. International joint conference on artificial intelligence 3
- LUCAS, KANADE
- 1981
(Show Context)
Citation Context ...as possible across the image plane. We implement this by randomly sampling points on the plane, searching then around that point for a feature with enough brightness gradient (we use an SSD-type test =-=[17]). O-=-nce a new point-feature is found (one with enough contrast along two independent directions), a newslter (which we call a \sublter") is initialized based on the model (15). Its evolution is given... |

896 | Shape and motion from image streams under orthography: a factorization method
- Tomasi, Kanade
- 1992
(Show Context)
Citation Context ...ourse a batch approach that optimizes simultaneously on all frames will perform better! 2 Our work falls within the category of causal motion and structure estimation that has a long and rich history =-=[10, 7, 18, 4, 23, 19, 32, 9, 30, 24, 8, 12, 26, 11, 31, 34, 33, 13, 25, 16, 1, 2, 22, 35, 15]-=-. Thesrst attempts to prove stability of the schemes proposed are recent [21]. Thesrst attempts to handle occlusions in a causal scheme 3 came only a few years ago [19, 29]. Our approach is similar in... |

294 |
Affine structure from motion
- Koenderink, Doorn
- 1991
(Show Context)
Citation Context ...ourse a batch approach that optimizes simultaneously on all frames will perform better! 2 Our work falls within the category of causal motion and structure estimation that has a long and rich history =-=[10, 7, 18, 4, 23, 19, 32, 9, 30, 24, 8, 12, 26, 11, 31, 34, 33, 13, 25, 16, 1, 2, 22, 35, 15]-=-. Thesrst attempts to prove stability of the schemes proposed are recent [21]. Thesrst attempts to handle occlusions in a causal scheme 3 came only a few years ago [19, 29]. Our approach is similar in... |

256 | Recursive estimation of motion structure and focal length
- Azarbayejani, Pentland
- 1995
(Show Context)
Citation Context ...ourse a batch approach that optimizes simultaneously on all frames will perform better! 2 Our work falls within the category of causal motion and structure estimation that has a long and rich history =-=[10, 7, 18, 4, 23, 19, 32, 9, 30, 24, 8, 12, 26, 11, 31, 34, 33, 13, 25, 16, 1, 2, 22, 35, 15]-=-. Thesrst attempts to prove stability of the schemes proposed are recent [21]. Thesrst attempts to handle occlusions in a causal scheme 3 came only a few years ago [19, 29]. Our approach is similar in... |

248 | A Paraperspective Factorization Method for Shape and Motion Recovery
- Poelman, Kanade
- 1997
(Show Context)
Citation Context |

238 |
Determining three-dimensional motion and structure from optical flow generated by several moving objects
- Adiv
- 1985
(Show Context)
Citation Context |

213 | Kalman filter-based algorithm for estimating depth from image sequences
- Matthies, Kanade, et al.
- 1989
(Show Context)
Citation Context |

191 | Recovering 3D shape and motion from image streams using nonlinear least squares
- Szeliski, Kang
- 1994
(Show Context)
Citation Context |

130 | Optimal motion and structure estimation
- Weng, Ahuja, et al.
- 1993
(Show Context)
Citation Context |

115 |
Estimation of object motion parameters from noisy images
- Broida, Chellappa
- 1986
(Show Context)
Citation Context |

80 |
Dynamic monocular machine vision
- Dickmanns, Graefe
- 1988
(Show Context)
Citation Context |

58 |
Three-dimensional motion computation and object segmentation in a long sequence of stereo frames
- Zhang, Faugeras
- 1992
(Show Context)
Citation Context |

55 | Shape recovery from multiple views: A parallax based approach,” presented at ARPA Image Understanding Workshop
- Kumar, Anandan, et al.
- 1994
(Show Context)
Citation Context |

52 |
Simplifying motion and structure analysis using planar parallax and image warping
- Sawhney
- 1994
(Show Context)
Citation Context |

49 |
A Multi-frame Approach to Visual Motion Perception
- Spetsakis, Aloimonos
- 1991
(Show Context)
Citation Context |

43 |
Inference and stochastic processes
- Bartlett
- 1967
(Show Context)
Citation Context ...e blocks corresponding to T (t) and (t) equal to zero (a deterministic integrator). We choose the remaining parameters using standard statistical tests, such as the Cumulative Periodogram of Bartlett =-=-=-[3]. The idea is that the parameters in w are changed until the innovation process (t) : = y(t) h( ^ (t)) is as close as possible to being white. The periodogram is one of many ways to test the \white... |

40 | Recursive Affine Structure and Motion From Image Sequences
- McLauchlan, Reid, et al.
- 1994
(Show Context)
Citation Context ...ourse a batch approach that optimizes simultaneously on all frames will perform better! 2 Our work falls within the category of causal motion and structure estimation that has a long and rich history =-=[10,7,18,4,23,19,32,9,30,24,8,12,26,11,31,34,33,13,25,16,1,2,22,35,15]-=-. The first attempts to prove stability of the schemes proposed are recent [21]. The first attempts to handle occlusions in a causal scheme 3 came only a few years ago [19,29]. Our approach is similar... |

40 |
Motion from point matches using affine epipolar geometry
- Shapiro, Zisserman, et al.
(Show Context)
Citation Context ...ourse a batch approach that optimizes simultaneously on all frames will perform better! 2 Our work falls within the category of causal motion and structure estimation that has a long and rich history =-=[10,7,18,4,23,19,32,9,30,24,8,12,26,11,31,34,33,13,25,16,1,2,22,35,15]-=-. The first attempts to prove stability of the schemes proposed are recent [21]. The first attempts to handle occlusions in a causal scheme 3 came only a few years ago [19,29]. Our approach is similar... |

34 |
Three Dimensional Vision, a geometric viewpoint
- Faugeras
- 1993
(Show Context)
Citation Context |

33 | Reducing Structure From Motion, a General Framework for Dynamic Vision part 1: Modeling
- Soatto, Perona
- 1998
(Show Context)
Citation Context ...,31,34,33,13,25,16,1,2,22,35,15]. The first attempts to prove stability of the schemes proposed are recent [21]. The first attempts to handle occlusions in a causal scheme 3 came only a few years ago =-=[19,29]-=-. Our approach is similar in spirit to the work of Azarbayejani and Pentland [2], extended to handle occlusions and to give correct weighting to the measurements. The first part of this study [5] cont... |

31 |
Subspace methods for recovering rigid motion, II: Theory
- AD, Heeger
(Show Context)
Citation Context |

30 | Direct recovery of motion and shape in the general case by xation
- Taalebinezhaad
- 1992
(Show Context)
Citation Context |

27 |
Motion and structure causally integrated over time
- Chiuso, Favaro, et al.
(Show Context)
Citation Context ...onal images can be posed as a nonlinearsltering problem. We describe the implementation of an algorithm whose uniform observability, minimal realization and stability have been proven analytically in =-=[5]-=-. We discuss a scheme for handling occlusions, drift in the scale factor and tuning of theslter. We also present an extension to partially calibrated camera models and prove its observability. We repo... |

26 |
Dynamic motion vision
- Heel
- 1989
(Show Context)
Citation Context |

25 |
Estimation of three-dimensional motion of rigid objects from noisy observations
- Philip
- 1991
(Show Context)
Citation Context |

22 | Recursive multi-frame structure from motion incorporatingmotion error
- Oliensis, Inigo-Thomas
- 1992
(Show Context)
Citation Context |

21 | Gauge Invariance in Projective 3D Reconstruction
- McLauchlan
- 1999
(Show Context)
Citation Context ...ot review them here. 4 Observability in SFM has been addressedsrst in 1994 [6, 27] (see also [28] for a more complete account of these results). Observability is closely related to \gauge invariance&q=-=uot; [20]. is -=-a minimal realization of (6). The notation Log SO(3) (R) stands for such that R = e b and is computed by inverting Rodrigues' formula 5 . is called the \canonical representation" of R. Remark 1 N... |

20 |
Tracking facilitates 3-D motion estimation
- Fermfiller, Aloimonos
- 1992
(Show Context)
Citation Context |

17 |
Visual tracking of known 3dimensional object”, Int
- Gennery
- 1992
(Show Context)
Citation Context |

14 | 3-D structure from visual motion : modeling, representation and observability
- Soatto
- 2006
(Show Context)
Citation Context ...l ways of handling missing data in a batch approach: since they do not extend to causal processing, we do not review them here. 4 Observability in SFM has been addressedsrst in 1994 [6, 27] (see also =-=[28] for -=-a more complete account of these results). Observability is closely related to \gauge invariance" [20]. is a minimal realization of (6). The notation Log SO(3) (R) stands for such that R = e b an... |

13 | Observability/identifiability of rigid motion under perspective projection
- Soatto
- 1994
(Show Context)
Citation Context ...3 There are several ways of handling missing data in a batch approach: since they do not extend to causal processing, we do not review them here. 4 Observability in SFM has been addressedsrst in 1994 =-=[6, 27] (see-=- also [28] for a more complete account of these results). Observability is closely related to \gauge invariance" [20]. is a minimal realization of (6). The notation Log SO(3) (R) stands for such ... |

12 |
A Necessary and Sufficient Condition for the Perspective Observability Problem
- Dayawansa, Ghosh, et al.
- 1995
(Show Context)
Citation Context ...There are several ways of handling missing data in a batch approach: since they do not extend to causal processing, we do not review them here. 4 Observability in SFM has been addressed first in 1994 =-=[6,27]-=- (see also [28] for a more complete account of these results). Observability is closely related to “gauge invariance” [20]. (7)sis a minimal realization of (6). The notation Log SO(3)(R) stands for Ω ... |

9 |
ªMotion and Structure Estimation Using Long Sequence Motion Models,º
- Hu, Ahuja
- 1993
(Show Context)
Citation Context |

8 | Provably correct algorithms for multi-frame structure from motion
- Oliensis
- 1996
(Show Context)
Citation Context ...that has a long and rich history [10, 7, 18, 4, 23, 19, 32, 9, 30, 24, 8, 12, 26, 11, 31, 34, 33, 13, 25, 16, 1, 2, 22, 35, 15]. Thesrst attempts to prove stability of the schemes proposed are recent =-=[21]-=-. Thesrst attempts to handle occlusions in a causal scheme 3 came only a few years ago [19, 29]. Our approach is similar in spirit to the work of Azarbayejani and Pentland [2], extended to handle occl... |

7 |
Real-time 3-d motion and structure from point features: a front-end system for vision-based control and interaction
- Jin, Favaro, et al.
- 2000
(Show Context)
Citation Context ...a biased or an unstableslter. The second part, reported in this paper, is concerned with the implementation of a system working in real time on real scenes, which we have made available to the public =-=[14]-=-. 2 Realization In order to design asnite-dimensional approximation to the optimalslter, we need an observable realization of the original model 4 . In [5] we have proven the following claim. Corollar... |

7 |
Recursive ane structure and motion from image sequences
- McLauchlan, Reid, et al.
- 1994
(Show Context)
Citation Context |

4 |
Motion from point matches using ane epipolar geometry
- Shapiro, Zisserman, et al.
- 1994
(Show Context)
Citation Context |

2 |
A necessary and su cient condition for the perspective observability problem
- Dayawansa, Ghosh, et al.
- 1995
(Show Context)
Citation Context ...3 There are several ways of handling missing data in a batch approach: since they do not extend to causal processing, we do not review them here. 4 Observability in SFM has been addressedsrst in 1994 =-=[6, 27] (see-=- also [28] for a more complete account of these results). Observability is closely related to \gauge invariance" [20]. is a minimal realization of (6). The notation Log SO(3) (R) stands for such ... |

2 |
Reducing \structure from motion": a general framework for dynamic vision. part 1: modeling
- Soatto, Perona
- 1998
(Show Context)
Citation Context ...4, 33, 13, 25, 16, 1, 2, 22, 35, 15]. Thesrst attempts to prove stability of the schemes proposed are recent [21]. Thesrst attempts to handle occlusions in a causal scheme 3 came only a few years ago =-=[19, 29]-=-. Our approach is similar in spirit to the work of Azarbayejani and Pentland [2], extended to handle occlusions and to give correct weighting to the measurements. Thesrst part of this study [5] contai... |