.png){kind=icon}
.png){kind=icon}
Return
Single-light-source visible light positioning using dual position-sensitive detectors
.png){kind=icon}
PRE
AI
DOI：10.1088/1361-6501/ae5321.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Visible light positioning (VLP) has emerged as a prominent research focus in indoor localization due to its immunity to electromagnetic interference and low deployment cost. However, the reliance on multiple light sources in conventional VLP systems poses a significant challenge to achieving large-scale coverage. To address the inherent dependence on multiple illuminated light sources used as beacons in VLP and realize seamless localization in areas without light coverage in conventional VLP systems, this paper proposes a single-light-source VLP system based on dual position-sensitive detectors (PSDs) and a loosely coupled fusion framework with visual-inertial odometry (VIO). Unlike conventional multi-light-sources VLP systems, the proposed system establishes a closed-form projection relationship between a single light source and a compact dual-PSD receiver. Through bundle-adjustment-style nonlinear optimization of the geometric reprojection error, millimeter-level four-degree-of-freedom global localization is realized. The proposed system achieves high localization accuracy with a significantly reduced number of light sources. Experiments demonstrate an average positioning accuracy of 7.5 mm within a 1 m2 area. Furthermore, the proposed system achieves seamless and continuous localization in non-illuminated regions, by integrating with VIO through a loosely coupled fusion framework. More importantly, it enables globally drift-correctable, large-scale positioning by leveraging the absolute global coordinates from the dual-PSD system to effectively suppress accumulated VIO drift.
Keywords:
Visible light positioning
dual position-sensitive detectors
single-light-source system
visual-inertial odometry
bundle adjustment
Journal
IF:
3.4
Papers:
2.6K
Citations:
2.3W