The Multi-view Geometry of Parallel Cylinders
(2023) 22nd Scandinavian Conference on Image Analysis, SCIA 2023 In Lecture Notes in Computer Science 13886. p.482-499- Abstract
- In this paper we study structure from motion problems for parallel cylinders. Using sparse keypoint correspondences is an efficient (and standard) way to solve the structure from motion problem. However, point features are sometimes unavailable and they can be unstable over time and viewing conditions. Instead, we propose a framework based on silhouettes of quadric surfaces, with special emphasis on parallel cylinders. Such structures are quite common, e.g. trees, lampposts, pillars, and furniture legs. Traditionally, the projection of the center lines of such cylinders have been considered and used in computer vision. Here, we demonstrate that the apparent width of the cylinders also contains useful information for structure and motion... (More)
- In this paper we study structure from motion problems for parallel cylinders. Using sparse keypoint correspondences is an efficient (and standard) way to solve the structure from motion problem. However, point features are sometimes unavailable and they can be unstable over time and viewing conditions. Instead, we propose a framework based on silhouettes of quadric surfaces, with special emphasis on parallel cylinders. Such structures are quite common, e.g. trees, lampposts, pillars, and furniture legs. Traditionally, the projection of the center lines of such cylinders have been considered and used in computer vision. Here, we demonstrate that the apparent width of the cylinders also contains useful information for structure and motion estimation. We provide mathematical analysis of relative structure and relative motion tensors, which is used to develop a number of minimal solvers for simultaneously estimating camera pose and scene structure from silhouette lines of cylinders. These solvers can be used efficiently in robust estimation schemes, such as RANSAC. We use Sampson-approximation methods for efficient estimation using over-determined data and develop averaging techniques. We also perform synthetic accuracy and robustness tests and evaluate our methods on a number of real-world scenarios. (Less)
- Abstract (Swedish)
- In this paper we study structure from motion problems for parallel cylinders. Using sparse keypoint correspondences is an efficient (and standard) way to solve the structure from motion problem. However, point features are sometimes unavailable and they can be unstable over time and viewing conditions. Instead, we propose a framework based on silhouettes of quadric surfaces, with special emphasis on parallel cylinders. Such structures are quite common, e.g. trees, lampposts, pillars, and furniture legs. Traditionally, the projection of the center lines of such cylinders have been considered and used in computer vision. Here, we demonstrate that the apparent width of the cylinders also contains useful information for structure and motion... (More)
- In this paper we study structure from motion problems for parallel cylinders. Using sparse keypoint correspondences is an efficient (and standard) way to solve the structure from motion problem. However, point features are sometimes unavailable and they can be unstable over time and viewing conditions. Instead, we propose a framework based on silhouettes of quadric surfaces, with special emphasis on parallel cylinders. Such structures are quite common, e.g. trees, lampposts, pillars, and furniture legs. Traditionally, the projection of the center lines of such cylinders have been considered and used in computer vision. Here, we demonstrate that the apparent width of the cylinders also contains useful information for structure and motion estimation. We provide mathematical analysis of relative structure and relative motion tensors, which is used to develop a number of minimal solvers for simultaneously estimating camera pose and scene structure from silhouette lines of cylinders. These solvers can be used efficiently in robust estimation schemes, such as RANSAC. We use Sampson-approximation methods for efficient estimation using over-determined data and develop averaging techniques. We also perform synthetic accuracy and robustness tests and evaluate our methods on a number of real-world scenarios. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/67ac8d99-87d0-4951-89de-6ddab9c74e86
- author
- Tegler, Erik LU ; Engman, Johanna LU ; Gillsjö, David LU ; Flood, Gabrielle LU ; Larsson, Viktor LU ; Oskarsson, Magnus LU and Åström, Kalle LU
- organization
-
- Mathematics (Faculty of Engineering)
- eSSENCE: The e-Science Collaboration
- ELLIIT: the Linköping-Lund initiative on IT and mobile communication
- LTH Profile Area: AI and Digitalization
- Mathematical Imaging Group (research group)
- LU Profile Area: Light and Materials
- LU Profile Area: Proactive Ageing
- LU Profile Area: Natural and Artificial Cognition
- Stroke Imaging Research group (research group)
- LTH Profile Area: Engineering Health
- publishing date
- 2023
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- host publication
- Image Analysis : 23rd Scandinavian Conference, SCIA 2023, Sirkka, Finland, April 18–21, 2023, Proceedings - 23rd Scandinavian Conference, SCIA 2023, Sirkka, Finland, April 18–21, 2023, Proceedings
- series title
- Lecture Notes in Computer Science
- editor
- Gade, Rikke ; Felsberg, Michael and Kämäräinen, Joni-Kristian
- volume
- 13886
- pages
- 482 - 499
- publisher
- Springer
- conference name
- 22nd Scandinavian Conference on Image Analysis, SCIA 2023
- conference location
- Sirkka, Finland
- conference dates
- 2023-04-18 - 2023-04-21
- external identifiers
-
- scopus:85161441544
- ISSN
- 1611-3349
- 0302-9743
- ISBN
- 978-3-031-31437-7
- 978-3-031-31438-4
- DOI
- 10.1007/978-3-031-31438-4_32
- language
- English
- LU publication?
- yes
- id
- 67ac8d99-87d0-4951-89de-6ddab9c74e86
- date added to LUP
- 2023-04-28 13:40:24
- date last changed
- 2024-09-21 11:25:24
@inproceedings{67ac8d99-87d0-4951-89de-6ddab9c74e86, abstract = {{In this paper we study structure from motion problems for parallel cylinders. Using sparse keypoint correspondences is an efficient (and standard) way to solve the structure from motion problem. However, point features are sometimes unavailable and they can be unstable over time and viewing conditions. Instead, we propose a framework based on silhouettes of quadric surfaces, with special emphasis on parallel cylinders. Such structures are quite common, e.g. trees, lampposts, pillars, and furniture legs. Traditionally, the projection of the center lines of such cylinders have been considered and used in computer vision. Here, we demonstrate that the apparent width of the cylinders also contains useful information for structure and motion estimation. We provide mathematical analysis of relative structure and relative motion tensors, which is used to develop a number of minimal solvers for simultaneously estimating camera pose and scene structure from silhouette lines of cylinders. These solvers can be used efficiently in robust estimation schemes, such as RANSAC. We use Sampson-approximation methods for efficient estimation using over-determined data and develop averaging techniques. We also perform synthetic accuracy and robustness tests and evaluate our methods on a number of real-world scenarios.}}, author = {{Tegler, Erik and Engman, Johanna and Gillsjö, David and Flood, Gabrielle and Larsson, Viktor and Oskarsson, Magnus and Åström, Kalle}}, booktitle = {{Image Analysis : 23rd Scandinavian Conference, SCIA 2023, Sirkka, Finland, April 18–21, 2023, Proceedings}}, editor = {{Gade, Rikke and Felsberg, Michael and Kämäräinen, Joni-Kristian}}, isbn = {{978-3-031-31437-7}}, issn = {{1611-3349}}, language = {{eng}}, pages = {{482--499}}, publisher = {{Springer}}, series = {{Lecture Notes in Computer Science}}, title = {{The Multi-view Geometry of Parallel Cylinders}}, url = {{http://dx.doi.org/10.1007/978-3-031-31438-4_32}}, doi = {{10.1007/978-3-031-31438-4_32}}, volume = {{13886}}, year = {{2023}}, }