Real-Time Camera Tracking and 3D Reconstruction Using Signed Distance Functions
(2013) Robotics: Science and Systems (RSS) Conference 2013 9.- Abstract
- The ability to quickly acquire 3D models is an essential capability needed in many disciplines including robotics, computer vision, geodesy, and architecture. In this paper we present a novel method for real-time camera tracking and 3D
reconstruction of static indoor environments using an RGB-D sensor. We show that by representing the geometry with a signed distance function (SDF), the camera pose can be efficiently estimated by directly minimizing the error of the depth images on the SDF. As the SDF contains the distances to the surface for
each voxel, the pose optimization can be carried out extremely fast. By iteratively estimating the camera poses and integrating the RGB-D data in the voxel grid, a detailed... (More) - The ability to quickly acquire 3D models is an essential capability needed in many disciplines including robotics, computer vision, geodesy, and architecture. In this paper we present a novel method for real-time camera tracking and 3D
reconstruction of static indoor environments using an RGB-D sensor. We show that by representing the geometry with a signed distance function (SDF), the camera pose can be efficiently estimated by directly minimizing the error of the depth images on the SDF. As the SDF contains the distances to the surface for
each voxel, the pose optimization can be carried out extremely fast. By iteratively estimating the camera poses and integrating the RGB-D data in the voxel grid, a detailed reconstruction of an indoor environment can be achieved. We present reconstructions of several rooms using a hand-held sensor and from onboard an autonomous quadrocopter. Our extensive evaluation on publicly
available benchmark data shows that our approach is more accurate and robust than the iterated closest point algorithm (ICP) used by KinectFusion, and yields often a comparable accuracy at much higher speed to feature-based bundle adjustment methods such as RGB-D SLAM for up to medium-sized scenes. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/3917422
- author
- Bylow, Erik LU ; Sturm, Jürgen ; Kerl, Christian ; Kahl, Fredrik LU and Cremers, Daniel
- organization
- publishing date
- 2013
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- host publication
- Robotics: Science and Systems (RSS), Online Proceedings
- volume
- 9
- pages
- 8 pages
- publisher
- Robotics: Science and Systems
- conference name
- Robotics: Science and Systems (RSS) Conference 2013
- conference location
- Berlin, Germany
- conference dates
- 2013-06-24 - 2013-06-28
- ISSN
- 2330-765X
- ISBN
- 978-981-07-3937-9
- language
- English
- LU publication?
- yes
- id
- afb271b3-a00a-4e99-85ce-12c885c81626 (old id 3917422)
- alternative location
- http://www.roboticsproceedings.org/rss09/p35.pdf
- date added to LUP
- 2016-04-01 13:56:12
- date last changed
- 2018-11-21 20:21:25
@inproceedings{afb271b3-a00a-4e99-85ce-12c885c81626, abstract = {{The ability to quickly acquire 3D models is an essential capability needed in many disciplines including robotics, computer vision, geodesy, and architecture. In this paper we present a novel method for real-time camera tracking and 3D<br/><br> reconstruction of static indoor environments using an RGB-D sensor. We show that by representing the geometry with a signed distance function (SDF), the camera pose can be efficiently estimated by directly minimizing the error of the depth images on the SDF. As the SDF contains the distances to the surface for<br/><br> each voxel, the pose optimization can be carried out extremely fast. By iteratively estimating the camera poses and integrating the RGB-D data in the voxel grid, a detailed reconstruction of an indoor environment can be achieved. We present reconstructions of several rooms using a hand-held sensor and from onboard an autonomous quadrocopter. Our extensive evaluation on publicly<br/><br> available benchmark data shows that our approach is more accurate and robust than the iterated closest point algorithm (ICP) used by KinectFusion, and yields often a comparable accuracy at much higher speed to feature-based bundle adjustment methods such as RGB-D SLAM for up to medium-sized scenes.}}, author = {{Bylow, Erik and Sturm, Jürgen and Kerl, Christian and Kahl, Fredrik and Cremers, Daniel}}, booktitle = {{Robotics: Science and Systems (RSS), Online Proceedings}}, isbn = {{978-981-07-3937-9}}, issn = {{2330-765X}}, language = {{eng}}, publisher = {{Robotics: Science and Systems}}, title = {{Real-Time Camera Tracking and 3D Reconstruction Using Signed Distance Functions}}, url = {{https://lup.lub.lu.se/search/files/3674772/3917437.pdf}}, volume = {{9}}, year = {{2013}}, }