Motion and structure from two perspective views: algorithms, error analysis, and error estimation

Weng, Juyang; Huang, Thomas S.; Ahuja, Narendra

doi:10.1109/34.24779

Cited by 400 publications

(204 citation statements)

References 27 publications

Supporting

Mentioning

185

Contrasting

Unclassified

Order By: Relevance

“…To propagate the error through the eigenvalue decomposition, we use the result presented by Weng et al in [12]. The linear error term in α, the smallest eigenvector of C T C, is given by…”

Section: B Error Analysismentioning

confidence: 99%

See 1 more Smart Citation

Measurement errors in visual servoing

Kyrki

Kragić

Christensen

2004

IEEE International Conference on Robotics and Automation, 2004. Proceedings. ICRA '04. 2004

View full text Add to dashboard Cite

“…To propagate the error through the eigenvalue decomposition, we use the result presented by Weng et al in [12]. The linear error term in α, the smallest eigenvector of C T C, is given by…”

Section: B Error Analysismentioning

confidence: 99%

“…This, so called Orthogonal Procrustes problem, can be solved using SVD as suggested by Fiore, but we choose to solve the rotation using unit quaternions as presented by Weng et al in [12]. The solution involves another eigenvalue decomposition for the solution of a set of homogeneous linear equations.…”

Section: A Estimation Algorithmmentioning

confidence: 99%

Measurement errors in visual servoing

Kyrki

Kragić

Christensen

2004

IEEE International Conference on Robotics and Automation, 2004. Proceedings. ICRA '04. 2004

View full text Add to dashboard Cite

“…eters for the cylinder. The error in the to estimate is large because the cylinder is rotating around an axis nearly parallel to the image plane and, as pointed out in [WH1], a rotation about an axis parallel to the image plane is inherently difficult to distinguish from translation parallel to the image plane and perpendicular to the rotation axis (this also explains the error in ~). Note that the predicted trajectory point positions still differ from the actual positions by an average of less than the error threshold of 0.5 pixel.…”

Section: Methodsmentioning

confidence: 97%

Motion and structure factorization and segmentation of long multiple motion image sequences

Debrunner

Ahuja

1992

Computer Vision — ECCV'92

View full text Add to dashboard Cite

Abstract. This paper presents a computer algorithm which, given a dense temporal sequence of intensity images of multiple moving objects, will separate the images into regions showing distinct objects, and for those objects which are rotating, will calculate the three-dimensional structure and motion. The method integrates the segmentation of trajectories into subsets corresponding to different objects with the determination of the motion and structure of the objects. Trajectories are partitioned into groups corresponding to the different objects by fitting the trajectories from each group to a hierarchy of increasingly complex motion models. This grouping algorithm uses an efficient motion estimation algorithm based on the factorization of a measurement matrix into motion and structure components. Experiments are reported using two real image sequences of 50 frames each to test the algorithm.

show abstract

“…Using the quaternion notation our minimization problem is equivalent to solving (see Weng et al [216] …”

Section: B24 Scalementioning

confidence: 99%

Environment Learning for Indoor Mobile Robots

Andrade‐Cetto¹

2006

Springer Tracts in Advanced Robotics

View full text Add to dashboard Cite

This thesis focuses on the various aspects of autonomous environment learning for indoor service robots. Particularly, on landmark extraction from sensor data, autonomous map building, and robot localization.To univocally identify landmarks from sensor data, we study several landmark representations, and the mathematical foundation necessary to extract the features that build them from images and laser range data. The features extracted from just one sensor may not suffice in the invariant characterization of landmarks and objects, pushing for the combination of information from multiple sources. We present a new algorithm that fuses complementary information from two low level vision modules into coherent object models that can be tracked and learned in a mobile robotics context. Illumination conditions and occlusions are the most prominent artifacts that hinder data association in computer vision. By using photogrammetric and geometric constraints we restrict the search for landmark matches in successive images, and by locking our interest in one or a set of landmarks in the scene, we track those landmarks along successive frames, reducing considerably the data association problem. We concentrate on those tools from the geometry of multiple views that are relevant to the computation of initial landmark location estimates for coarse motion recovery; a desirable characteristic when odometry is not available or is highly unreliable.Once landmarks are accurately extracted and identified, the second part of the problem is to use these observations for the localization of the robot, as well as the refinement of the landmark location estimates. We consider robot motion and sensor observations as stochastic processes, and treat the problem from an estimation theoretic point of view, dealing with noise by using probabilistic methods.The main drawback we encounter is that current estimation techniques have been devised for static environments, and that they lack robustness in more realistic situations. To aid in those situations in which landmark observations might not be consistent in time, we propose a new set of temporal landmark quality functions, and show how iii by incorporating these functions in the data association tests, the overall estimationtheoretic approach to map building and localization is improved. The basic idea consists on using the history of data association mismatches for the computation of the likelihood of future data association, together with the spatial compatibility tests already available.Special attention is paid in that the removal of spurious landmarks from the map does not violate the basic convergence properties of the localization and map building algorithms already described in the literature; namely, asymptotic convergence and full correlation.The thesis also gives an in depth analysis of the fully correlated model to localization and map building from a control systems theory point of view. Considering the fact that the Kalman filter is nothing else but an optimal observer,...

show abstract

Motion and structure from two perspective views: algorithms, error analysis, and error estimation

Cited by 400 publications

References 27 publications

Measurement errors in visual servoing

Measurement errors in visual servoing

Motion and structure factorization and segmentation of long multiple motion image sequences

Environment Learning for Indoor Mobile Robots

Contact Info

Product

Resources

About