Homography-based Egomotion Estimation Using Gravity and SIFT Features
(2020) 15th Asian Conference on Computer Vision, ACCV 2020 In Lecture Notes in Computer Science 12622. p.278-294- Abstract
- Camera systems used, e.g., in cars, UAVs, smartphones, and tablets, are typically equipped with IMUs (inertial measurement units) that can measure the gravity vector. Using the information from an IMU, the y-axes of cameras can be aligned with the gravity, reducing their relative orientation to a single DOF (degree of freedom). In this paper, we use the gravity information to derive extremely efficient minimal solvers for homography-based egomotion estimation from orientation- and scale-covariant features. We use the fact that orientation- and scale-covariant features, such as SIFT or ORB, provide additional constraints on the homography. Based on the prior knowledge about the target plane (horizontal/vertical/general plane, w.r.t. the... (More)
- Camera systems used, e.g., in cars, UAVs, smartphones, and tablets, are typically equipped with IMUs (inertial measurement units) that can measure the gravity vector. Using the information from an IMU, the y-axes of cameras can be aligned with the gravity, reducing their relative orientation to a single DOF (degree of freedom). In this paper, we use the gravity information to derive extremely efficient minimal solvers for homography-based egomotion estimation from orientation- and scale-covariant features. We use the fact that orientation- and scale-covariant features, such as SIFT or ORB, provide additional constraints on the homography. Based on the prior knowledge about the target plane (horizontal/vertical/general plane, w.r.t. the gravity direction) and using the SIFT/ORB constraints, we derive new minimal solvers that require fewer correspondences than traditional approaches and, thus, speed up the robust estimation procedure significantly. The proposed solvers are compared with the state-of-the-art point-based solvers on both synthetic data and real images, showing comparable accuracy and significant improvement in terms of speed. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/f43e9581-e2ae-4113-ba0a-c886a734708c
- author
- Ding, Yaqing LU ; Barath, Daniel and Kukelova, Zuzana
- publishing date
- 2020
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- host publication
- Computer Vision – ACCV 2020 : 15th Asian Conference on Computer Vision, Kyoto, Japan, November 30 – December 4, 2020, Revised Selected Papers, Part I - 15th Asian Conference on Computer Vision, Kyoto, Japan, November 30 – December 4, 2020, Revised Selected Papers, Part I
- series title
- Lecture Notes in Computer Science
- volume
- 12622
- pages
- 278 - 294
- publisher
- Springer
- conference name
- 15th Asian Conference on Computer Vision, ACCV 2020
- conference location
- Virtual, Online
- conference dates
- 2020-11-30 - 2020-12-04
- external identifiers
-
- scopus:85103250725
- ISSN
- 1611-3349
- 0302-9743
- ISBN
- 978-3-030-69524-8
- 978-3-030-69525-5
- DOI
- 10.1007/978-3-030-69525-5_17
- language
- English
- LU publication?
- no
- id
- f43e9581-e2ae-4113-ba0a-c886a734708c
- date added to LUP
- 2022-09-09 10:57:10
- date last changed
- 2024-03-21 04:34:34
@inbook{f43e9581-e2ae-4113-ba0a-c886a734708c, abstract = {{Camera systems used, e.g., in cars, UAVs, smartphones, and tablets, are typically equipped with IMUs (inertial measurement units) that can measure the gravity vector. Using the information from an IMU, the y-axes of cameras can be aligned with the gravity, reducing their relative orientation to a single DOF (degree of freedom). In this paper, we use the gravity information to derive extremely efficient minimal solvers for homography-based egomotion estimation from orientation- and scale-covariant features. We use the fact that orientation- and scale-covariant features, such as SIFT or ORB, provide additional constraints on the homography. Based on the prior knowledge about the target plane (horizontal/vertical/general plane, w.r.t. the gravity direction) and using the SIFT/ORB constraints, we derive new minimal solvers that require fewer correspondences than traditional approaches and, thus, speed up the robust estimation procedure significantly. The proposed solvers are compared with the state-of-the-art point-based solvers on both synthetic data and real images, showing comparable accuracy and significant improvement in terms of speed.}}, author = {{Ding, Yaqing and Barath, Daniel and Kukelova, Zuzana}}, booktitle = {{Computer Vision – ACCV 2020 : 15th Asian Conference on Computer Vision, Kyoto, Japan, November 30 – December 4, 2020, Revised Selected Papers, Part I}}, isbn = {{978-3-030-69524-8}}, issn = {{1611-3349}}, language = {{eng}}, pages = {{278--294}}, publisher = {{Springer}}, series = {{Lecture Notes in Computer Science}}, title = {{Homography-based Egomotion Estimation Using Gravity and SIFT Features}}, url = {{http://dx.doi.org/10.1007/978-3-030-69525-5_17}}, doi = {{10.1007/978-3-030-69525-5_17}}, volume = {{12622}}, year = {{2020}}, }