Global Motion Estimation from Point Matches

Arie-Nachimson, Mica; Kovalsky, Shahar Z.; Kemelmacher-Shlizerman, Ira; Singer, Amit; Basri, Ronen

doi:10.1109/3dimpvt.2012.46

Cited by 178 publications

(338 citation statements)

References 28 publications

Supporting

Mentioning

338

Contrasting

Order By: Relevance

“…in Arie-Nachimson et al (2012). It is important to note that (6) is not robust against outliers; we argue that all defective rotations have already been eliminated during graph optimization.…”

Section: So(3) Relaxed Optimizationmentioning

confidence: 96%

Convex Image Orientation From Relative Orientations

Reich

Heipke

2016

ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:In this paper we propose a novel workflow for the estimation of global image orientations given relative orientations between pairs of overlapping images. Our approach is convex and independent on initial values. First, global rotations are estimated in a relaxed semidefinite program (SDP) and refined in an iterative least squares adjustment in the tangent space of SO(3). A critical aspect is the handling of outliers in the relative orientations. We present a novel heuristic graph based approach for filtering the relative rotations that outperforms state-of-the-art robust rotation averaging algorithms. In a second part we make use of point-observations, tracked over a set of overlapping images and formulate a linear homogeneous system of equations to transfer the scale information between triplets of images, using estimated global rotations and relative translation directions. The final step consists of refining the orientation parameters in a robust bundle adjustment. The proposed approach handles outliers in the homologous points and relative orientations in every step of the processing chain. We demonstrate the robustness of the procedure on synthetic data. Moreover, the performance of our approach is illustrated on real world benchmark data.

show abstract

Section: So(3) Relaxed Optimizationmentioning

confidence: 96%

Convex Image Orientation From Relative Orientations

Reich

Heipke

2016

ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

show abstract

“…In (Martinec & Pajdla, 2007) the constraints are enforced in a subsequent step. (Arie-Nachimson et al, 2012) include these constraints while using semi-definite programing. The approach is adopted and enhanced in (Reich & Heipke, 2016).…”

Section: Related Workmentioning

confidence: 99%

“…The latter suggest a reduction of the number of reconstructed points by selecting four representative points per image pair to reduce the computational workload. (Arie-Nachimson et al, 2012) makes use of point observations, but reformulates the epipolar constraint by replacing relative orientations with absolute ones. With fixed rotations, the resulting system is linear and can be solved efficiently.…”

Section: Related Workmentioning

confidence: 99%

Hierarchical Structure From Motion Combining Global Image Orientation and Structureless Bundle Adjustment

Cefalu

Haala

Fritsch

2017

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:Global image orientation techniques aim at estimating camera rotations and positions for a whole set of images simultaneously. One of the main arguments for these procedures is an improved robustness against drifting of camera stations in comparison to more classical sequential approaches. Usually, the process consists of computation of absolute rotations and, in a second step, absolute positions for the cameras. Either the first or both steps rely on the network of transformations arising from relative orientations between cameras. Therefore, the quality of the obtained absolute results is influenced by tensions in the network. These may e.g. be induced by insufficient knowledge of the intrinsic camera parameters. Another reason can be found in local weaknesses of image connectivity. We apply a hierarchical approach with intermediate bundle adjustment to reduce these effects. We adopt efficient global techniques which register image triplets based on fixed absolute camera rotations and scaled relative camera translations but do not involve scene structure elements in the fusion step. Our variant employs submodels of arbitrary size, orientation and scale, by computing relative rotations and scales between -and subsequently absolute rotations and scales for -submodels and is applied hierarchically. Furthermore we substitute classical bundle adjustment by a structureless approach based on epipolar geometry and augmented with a scale consistency constraint.

show abstract

“…Secondly, we formulate a novel algorithm to remove the outliers among the relative rotations, presented in Section 5. Finally, in Section 6 we provide a theoretical analysis of the linear algorithm introduced in (Arie-Nachimson et al, 2012), that solve for exterior positions. The discussion carried out in the paper is supported by experimental results on both synthetic and real images, shown in Section 7.…”

Section: Overviewmentioning

confidence: 99%

“…The advantage of this expression is that pairwise information is no longer required. As explained in (Arie-Nachimson et al, 2012), the epipolar constraint defined by (18) leads to a linear equation for every pair of matching points…”

Section: Position Recovery: a New Interpretationmentioning

confidence: 99%

Robust Global Motion Estimation with Matrix Completion

Arrigoni

Rossi

Malapelle

et al. 2014

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:In this paper we address the problem of estimating the attitudes and positions of a set of cameras in an external coordinate system. Starting from a conventional global structure-from-motion pipeline, we present some substantial advances. In order to detect outlier relative rotations extracted from pairs of views, we improve a state-of-the-art algorithm based on cycle consistency, by introducing cycle bases. We estimate the angular attitudes of the cameras by proposing a novel gradient descent algorithm based on low-rank matrix completion, that naturally copes with the case of missing data. As for position recovery, we analyze an existing technique from a theoretical point of view, providing some insights on the conditions that guarantee solvability. We provide experimental results on both synthetic and real image sequences for which ground truth calibration is provided.

show abstract

Global Motion Estimation from Point Matches

Cited by 178 publications

References 28 publications

Convex Image Orientation From Relative Orientations

Convex Image Orientation From Relative Orientations

Hierarchical Structure From Motion Combining Global Image Orientation and Structureless Bundle Adjustment

Robust Global Motion Estimation with Matrix Completion

Contact Info

Product

Resources

About