ElasticFusion: real-time dense SLAM and light source estimation
- Submitting institution
-
Imperial College of Science, Technology and Medicine
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 2240
- Type
- D - Journal article
- DOI
-
10.1177/0278364916669237
- Title of journal
- International Journal of Robotics Research
- Article number
- 14
- First page
- 1697
- Volume
- 35
- Issue
- 14
- ISSN
- 1741-3176
- Open access status
- Compliant
- Month of publication
- September
- Year of publication
- 2016
- URL
-
-
- Supplementary information
-
10.1177/0278364916669237
- 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
- 186
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The approach constitutes a paradigm shift away from pose-graph optimisation to a map-centric dense SLAM system. It also includes an approach for estimating light sources and showing increased accuracy when considering related specularities. The paper is an invited extension of a highly-cited RSS 2015 paper (http://roboticsproceedings.org/rss11/p01.pdf). Our open-source implementation (https://bitbucket.org/dysonroboticslab/elasticfusionpublic) is widely adopted and used for comparisons to newer approaches (e.g., https://bit.ly/3cddxkr, https://bit.ly/3ooqmLp, https://bit.ly/3ond3uC, https://bit.ly/36j9ejR, https://bit.ly/36jJ0xA). It has also been used internally for the development of the dense semantic mapping system, SemanticFusion, as well as externally, e.g., MaskFusion (http://visual.cs.ucl.ac.uk/pubs/maskfusion/) and CoFusion (http://visual.cs.ucl.ac.uk/pubs/cofusion/).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
