2011 International Conference on Computer Vision 2011
DOI: 10.1109/iccv.2011.6126530
|View full text |Cite
|
Sign up to set email alerts
|

Optimal estimation of vanishing points in a Manhattan world

Abstract: In this paper, we present an analytical method for computing the globally optimal estimates of orthogonal vanishing points in a "Manhattan world" with a calibrated camera. We formulate this as constrained least-squares problem whose optimality conditions form a multivariate polynomial system. We solve this system analytically to compute all the critical points of the least-squares cost function, and hence the global minimum, i.e., the globally optimal estimate for the orthogonal vanishing points. The same opti… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
66
0

Year Published

2013
2013
2023
2023

Publication Types

Select...
5
3

Relationship

0
8

Authors

Journals

citations
Cited by 87 publications
(66 citation statements)
references
References 19 publications
0
66
0
Order By: Relevance
“…This approach is used by Censi [13] for the global resolution of the 2D registration problem, reducing the problem to solving a 4-th order polynomial equation. However, this approach does not generalize well to the 3D case due to the higher complexity of the rotation space, which produces an explosion in the complexity of the resulting polynomial system [29,30].…”
Section: Global Optimizationmentioning
confidence: 99%
“…This approach is used by Censi [13] for the global resolution of the 2D registration problem, reducing the problem to solving a 4-th order polynomial equation. However, this approach does not generalize well to the 3D case due to the higher complexity of the rotation space, which produces an explosion in the complexity of the resulting polynomial system [29,30].…”
Section: Global Optimizationmentioning
confidence: 99%
“…Column [9] should be omitted from fair comparison because the authors seem to apply the tolerance on individual angles (pitch, yaw, roll), not on the overall angular difference, which is the case of other works (and our evaluation). Column [15] should also be treated lightly because the authors feed the detector only with edges user-annotated as belonging to one of the vanishing points (no outliers), while other works are working with all the edges in the image, including clutter. evaluation was done on the complete set in order to be comparable with previous works.…”
Section: Vanishing Point Detection Accuracymentioning
confidence: 99%
“…evaluation was done on the complete set in order to be comparable with previous works. We used two means of evaluation popular in the literature: detection rate with 10 • angular error tolerance [9,10,15,16,17,18], Table 1, and cumulative histogram of the count of correctly recognized vanishing points based on the angular error tolerance [2,13,14], Figure 7.…”
Section: Vanishing Point Detection Accuracymentioning
confidence: 99%
See 1 more Smart Citation
“…Firstly, we employ a RANSAC-based vanishing point estimator that uses triplets of lines for generating hypotheses of all three orthogonal vanishing points at once [13]. We prune these hypotheses, by keeping the one that corresponds to a rotational matrix with roll and pitch angles closest to those estimated by the filter.…”
Section: A Yaw Initializationmentioning
confidence: 99%