测绘通报 ›› 2019, Vol. 0 ›› Issue (5): 25-29.doi: 10.13474/j.cnki.11-2246.2019.0143

• 室内定位与导航 • 上一篇    下一篇

基于奇异值分解的快速室内可见光定位算法

肖志良1,2   

  1. 1. 佛山职业技术学院电子信息学院, 广东 佛山 528137;
    2. 武汉大学信息管理学院, 湖北 武汉 430072
  • 收稿日期:2018-08-20 修回日期:2018-11-30 出版日期:2019-05-25 发布日期:2019-06-04
  • 作者简介:肖志良(1965-),男,博士,副教授,高级工程师,研究方向为定位算法及物联网系统等。E-mail:zlfours@126.com

A fast indoor visible light positioning algorithm based on singular value decomposition

XIAO Zhiliang1,2   

  1. 1. College of Electronic Information, Foshan Polytechnic, Foshan 528137, China;
    2. College of Information Management, Wuhan University, Wuhan 430072, China
  • Received:2018-08-20 Revised:2018-11-30 Online:2019-05-25 Published:2019-06-04

摘要: 为了提高室内可见光定位的稳健性和效率,对定位系统的算法部分进行优化,将定位问题考虑为三维点位拟合问题,利用奇异值分解技术,推导出一个用于确定接收器位置和朝向的闭合表达式。另外,使用智能手机的图像传感器作为接收器,对平面部署的35个网格点进行了测试。与常用的LM迭代式定位算法相比,本文算法定位速度更快,大约快30倍;且避免了由于错误的初始猜测而造成定位失败的可能性,在实际环境中实现了厘米级的高精度定位。

关键词: 可见光定位, 奇异值分解, 拟合问题, 图像传感器, 接收器

Abstract: In order to improve the robustness and efficiency of indoor visible light positioning, the algorithm part of the positioning system is optimized. The positioning problem is considered as a three-dimensional point-to-point fitting problem. By using singular value decomposition (SVD) technique, a closed-form expression is derived to determine the position and orientation of the receiver. In addition, the image sensor of the smart phone is used as a receiver. 35 grid points are tested for the plane deployment. Compared with the commonly used LM iterative positioning algorithm, the proposed algorithm has faster positioning speed, about 30 times faster. Meanwhile, the proposed algorithm has avoided the possibility of positioning failure caused by wrong initial guess, and it achieves centimeter-level high-precision positioning in the actual environment.

Key words: visible light positioning, singular value decomposition, fitting problem, image sensor, receiver

中图分类号: