Research On Indoor Positioning Method Based On Visible Light Communication

With the development of mobile Internet technology,a variety of location-based services become more and more popular,such as location-based information pushing to provide shopping,entertainment,traveling and navigation services.On the other hand,people in modern society spend more than 80% of the time indoors according to statistics,such as shopping malls,museums,libraries and other large indoor spaces,and indoor positioning becomes an important demand.Indoor positioning usually requires a higher positioning accuracy,generally at least in meter level,and there is not a mature technical standard to be widely used yet.This paper first analyzes the current indoor positioning technologies,such as Wi-Fi,infrared,ultrasonic,ultra-wideband,RFID and so on,pointing out the major deficiencies of low accuracy and infrastructure complexity,and then proposes a VLC-based indoor positioning system.This kind of visible light localization method is quite different from the traditional RF-based positioning;because of the visible light transmission characteristics,the signal spreading in a relatively restricted range,positioning accuracy is greatly improved.Also,visible light communication uses the off the shelf lighting infrastructure,which is more cost-efficient.The paper first analyzes the principle of visible light localization method.It can be divided into two categories based on two kinds of receivers: non-imaging localization based on photodetector receiver and imaging localization based on camera receiver.The imaging method is adopted in this paper,which uses the common camera equipments such as smartphones as receivers.And the positioning accuracy is higher also the application prospect is better than the non-imaging method.The camera-based method involves two main processes.Firstly,the visible light communication between the camera and the light is realized to get the coordinate information from the luminaire.Secondly,the imaging algorithm is used to obtain the final position.By leveraging the CMOS camera rolling-shutter effect,the flashing light can cause banding effect on the image.Through an image processing procedure,frequency can be detected from the width of band,and information can be decoded definitely.Based on the imaging geometry of the camera,a positioning algorithm is proposed.It is needed to capture at least three light sources on the image and obtain their actual spatial coordinates and image coordinates,and finally a two-dimensional positioning can be realized.Then,the implementation process of visible light communication is introduced.The scheme uses a kind of household LED luminaire and embeds it with a modulation module,so that it can send frequency signals,as the transmitter.And a camera is used as the receiver to capture images for image processing.The paper uses OpenCV functions to achieve multi-light source extraction,and determine width of bands to detect the light frequency.Also,a modulation scheme and a communication protocol are designed to achieve a stable communication process.Finally,the implement of positioning system is introduced.An experimental testbed of 50 cm * 50 cm is set up,and more than three light sources with certain locations are put on it as reference points.The paper uses a 2 million pixel USB camera as a receiver,and designs the positioning software based on PC to display real-time positioning results.Positioning performance based on conditions of different numbers of luminaires has been tested and the results show that the error can be limited within 5cm under the coverage area of lights,which proves the superiority of the visible light positioning method.