Keyframe-based visual–inertial odometry using nonlinear optimization
- Submitting institution
-
Imperial College of Science, Technology and Medicine
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 2409
- Type
- D - Journal article
- DOI
-
10.1177/0278364914554813
- Title of journal
- The International Journal of Robotics Research
- Article number
- -
- First page
- 314
- Volume
- 34
- Issue
- 3
- ISSN
- 0278-3649
- Open access status
- Out of scope for open access requirements
- Month of publication
- December
- Year of publication
- 2014
- URL
-
-
- Supplementary information
-
10.1177/0278364914554813
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
4
- Research group(s)
-
-
- Citation count
- 494
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The approach expands a RSS 2013 paper (http://roboticsproceedings.org/rss09/p37.pdf) in describing and extensively evaluating a visual-inertial odometry (VIO) method. It underpins the open-source release of Open Keyframe-based Visual-Inertial SLAM (OKVIS), the localisation system on our drones, since adopted by the wider robotics community (https://github.com/ethz-asl/okvis; >300 forks). Our approach is used as a baseline for comparison by many VIO papers and forms the basis for the motion-tracking used by start-up SLAMcore (www.slamcore.com), co-founded by Leutenegger. It received Venture Capital in 2017 and 2018 and currently employs 15 people.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
