.png){kind=icon}
.png){kind=icon}
Return
Enforcing Regularities between Planes Using Key Plane for Monocular Mesh-based VIO
.png){kind=icon}
PRE
AI
DOI：10.1007/s10846-021-01529-5.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Enforcing the additional structural constraints to improve localization, the plane feature can express better environment information, and thereby is widely used in Visual Inertial Odometry (VIO) method. To enforce regularities between planes, this paper proposes a novel factor graph formulation based on key plane. Firstly, key plane selection criterion is adopted to filter trivial planes. Then structural regularities between each plane detected in previous steps and key plane are implemented by adding corresponding residual terms into optimization function. Furthermore, in addition to gravity, the normal directions of all key planes continually update actual buildings plane normal direction in the world frame, which is used for detecting plane. Finally, experiments was conducted on public dataset and results show that our proposed method obtains better accuracy than state-of-the-art algorithm.
Keywords:
VIO
Key plane feature
Structural regularities
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
J
IF:
2.8
Papers:
3.8K
Citations:
6.9K