2018
DOI: 10.1016/j.cviu.2018.09.007
|View full text |Cite
|
Sign up to set email alerts
|

Surface reconstruction from a sparse point cloud by enforcing visibility consistency and topology constraints

Abstract: Highlights• Surface reconstruction under-uses topology constraints in Computer Vision• Start from a previous method enforcing manifoldness on a sparse point cloud• Simultaneously enforce visibility consistency and low genus for the first time• Improve the removal of surface singularities and the escape from local extrema• Experiment on long video sequences taken by a helmetheld omnidirectional camera AbstractThere are reasons to reconstruct a surface from a sparse cloud of 3D points estimated from an image seq… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

0
19
0

Year Published

2019
2019
2023
2023

Publication Types

Select...
4
3

Relationship

2
5

Authors

Journals

citations
Cited by 15 publications
(19 citation statements)
references
References 39 publications
0
19
0
Order By: Relevance
“…However, these are only used if the number of points m is not too different from the number of points n. When there is a big difference between m and n, the regions of the point set with fewer points correspond to multiple edges on the point set with more points, and the constructed 3D structure becomes rough, and, in some instances, even loses the characteristics of the structure itself [8][9][10][11][12]. Another method is to use the Delaunay irregular triangulation mesh to realize the segmentation and modeling of discrete data points in space [13][14][15]. For a uniform set of spatial points, the technology of using Delaunay triangulation to form a spatial surface is relatively mature.…”
Section: Problems In Current Surface Reconstruction Technologiesmentioning
confidence: 99%
See 1 more Smart Citation
“…However, these are only used if the number of points m is not too different from the number of points n. When there is a big difference between m and n, the regions of the point set with fewer points correspond to multiple edges on the point set with more points, and the constructed 3D structure becomes rough, and, in some instances, even loses the characteristics of the structure itself [8][9][10][11][12]. Another method is to use the Delaunay irregular triangulation mesh to realize the segmentation and modeling of discrete data points in space [13][14][15]. For a uniform set of spatial points, the technology of using Delaunay triangulation to form a spatial surface is relatively mature.…”
Section: Problems In Current Surface Reconstruction Technologiesmentioning
confidence: 99%
“…For example, in real wellbore surface scan data, due to the fast drilling process, the data points are sometimes dense, and other times sparse and uneven. When the set of points on the borehole wall is very sparse, the Delaunay subdivision may treat the sparse points on the Another method is to use the Delaunay irregular triangulation mesh to realize the segmentation and modeling of discrete data points in space [13][14][15]. For a uniform set of spatial points, the technology of using Delaunay triangulation to form a spatial surface is relatively mature.…”
Section: Problems In Current Surface Reconstruction Technologiesmentioning
confidence: 99%
“…The latter tries to avoid triangle that is face of "too flat" tetrahedron. In [10], the surface is computed as the boundary of an evolving set of tetrahedra that is maintained manifold: the set mostly grows in the tetrahedra crossed by lines-of-sight to meet the visibility constraint. Our experiments show that our corrections based on LC improve the results of both [18] and [10].…”
Section: Previous Workmentioning
confidence: 99%
“…In [10], the surface is computed as the boundary of an evolving set of tetrahedra that is maintained manifold: the set mostly grows in the tetrahedra crossed by lines-of-sight to meet the visibility constraint. Our experiments show that our corrections based on LC improve the results of both [18] and [10]. Our corrections can also be used with most surface reconstruction methods, which are based on graph-cut optimization or manifold constraint.…”
Section: Previous Workmentioning
confidence: 99%
“…The neural network service would provide the exact floor level, while the camera collected data would be automatically compared with information stored on Google Maps to determine the precise coordinates. In recent years, complex surface reconstruction algorithms that enhance structure recognition have been developed and applied in VR applications [19].…”
Section: Proposed Software Solutionmentioning
confidence: 99%