Affiliation of Author(s):计算机科学与技术学院/人工智能学院/软件学院
Journal:IEEE PHOTONICS JOURNAL
Key Words:Visible light positioning computer vision inertial measurement unit image sensor
Abstract:Visible light positioning (VLP) has developed rapidly with the ubiquitous lighting infrastructure in recent years. Based on classic trilateration or triangulation, VLP usually requires at least three light sources which makes it infeasible in some scenarios where sparse LED luminaries are mounted. In this paper, we focus on VLP in such scenarios and propose a VLP method based on a single LED luminaire as the light transmitter and commercial off-the-shelf smartphone as the receiver. A critical problem that needs to be addressed is to derive the accurate orientation of the smartphone. Although the inertial measurement unit of the smartphone can be used to obtain the orientation, the measurement error severely degrades the positioning accuracy. To address the measurement error, this paper innovates in exploiting projective geometry to calibrate the orientation measurements. We build an experimental platform to evaluate the proposed VLP method. The experimental results show that the proposed VLP method can achieve a submeter accuracy and the indoor positioning errors are within 0.16 m by 90%.
ISSN No.:1943-0655
Translation or Not:no
Date of Publication:2019-10-01
Co-author:W00118,F70206470
Correspondence Author:haojie
Date of Publication:2019-10-01