Tracking 3-D Rotations with the Quaternion Bingham Filter
A deterministic method for sequential estimation of 3-D rotations is presented. The Bingham distribution is used to represent uncertainty directly on the unit quaternion hypersphere. Quaternions avoid the degeneracies of other 3-D orientation representations, while the Bingham distribution allows tr...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Published: |
2013
|
| Online Access: | http://hdl.handle.net/1721.1/78248 |
| _version_ | 1826207758748221440 |
|---|---|
| author | Glover, Jared Kaelbling, Leslie Pack |
| author2 | Leslie Kaelbling |
| author_facet | Leslie Kaelbling Glover, Jared Kaelbling, Leslie Pack |
| author_sort | Glover, Jared |
| collection | MIT |
| description | A deterministic method for sequential estimation of 3-D rotations is presented. The Bingham distribution is used to represent uncertainty directly on the unit quaternion hypersphere. Quaternions avoid the degeneracies of other 3-D orientation representations, while the Bingham distribution allows tracking of large-error (high-entropy) rotational distributions. Experimental comparison to a leading EKF-based filtering approach on both synthetic signals and a ball-tracking dataset shows that the Quaternion Bingham Filter (QBF) has lower tracking error than the EKF, particularly when the state is highly dynamic. We present two versions of the QBF, suitable for tracking the state of first- and second-order rotating dynamical systems. |
| first_indexed | 2024-09-23T13:54:30Z |
| id | mit-1721.1/78248 |
| institution | Massachusetts Institute of Technology |
| last_indexed | 2024-09-23T13:54:30Z |
| publishDate | 2013 |
| record_format | dspace |
| spelling | mit-1721.1/782482019-04-10T23:14:39Z Tracking 3-D Rotations with the Quaternion Bingham Filter Glover, Jared Kaelbling, Leslie Pack Leslie Kaelbling Learning and Intelligent Systems A deterministic method for sequential estimation of 3-D rotations is presented. The Bingham distribution is used to represent uncertainty directly on the unit quaternion hypersphere. Quaternions avoid the degeneracies of other 3-D orientation representations, while the Bingham distribution allows tracking of large-error (high-entropy) rotational distributions. Experimental comparison to a leading EKF-based filtering approach on both synthetic signals and a ball-tracking dataset shows that the Quaternion Bingham Filter (QBF) has lower tracking error than the EKF, particularly when the state is highly dynamic. We present two versions of the QBF, suitable for tracking the state of first- and second-order rotating dynamical systems. 2013-03-29T21:30:06Z 2013-03-29T21:30:06Z 2013-03-27 http://hdl.handle.net/1721.1/78248 MIT-CSAIL-TR-2013-005 Creative Commons Attribution 3.0 Unported http://creativecommons.org/licenses/by/3.0/ 13 p. application/pdf |
| spellingShingle | Glover, Jared Kaelbling, Leslie Pack Tracking 3-D Rotations with the Quaternion Bingham Filter |
| title | Tracking 3-D Rotations with the Quaternion Bingham Filter |
| title_full | Tracking 3-D Rotations with the Quaternion Bingham Filter |
| title_fullStr | Tracking 3-D Rotations with the Quaternion Bingham Filter |
| title_full_unstemmed | Tracking 3-D Rotations with the Quaternion Bingham Filter |
| title_short | Tracking 3-D Rotations with the Quaternion Bingham Filter |
| title_sort | tracking 3 d rotations with the quaternion bingham filter |
| url | http://hdl.handle.net/1721.1/78248 |
| work_keys_str_mv | AT gloverjared tracking3drotationswiththequaternionbinghamfilter AT kaelblinglesliepack tracking3drotationswiththequaternionbinghamfilter |