Abstract | Keywords | Export | Availability | Bookmark

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

With the boom of the internet of things and the next generation of the automation industry, along with the imagination and expectation of convenient and intelligent future life, the demand for high-accuracy, low-latency indoor positioning systems is increasing day by day. However, the traditional outdoor positioning system cannot effectively cope with complex indoor situations. The low indoor accuracy made researchers looking for different proposals for indoor positioning. Therefore, indoor positioning systems have mushroomed. Many methods utilizing different technologies have been proposed. Wi-Fi, Bluetooth, Ultra-Wide Band, ultrasound, infrared, visible light, and other technologies all see their application in the field of indoor positioning. Among all the competitors, Visible Light Positioning (VLP), which typically uses LEDs as transmitters, stands out for its high accuracy and low-cost hardware. For precise positioning, the accurate position of all LEDs must be known in advance. However, in reality, these conditions may not all be met. Thus, a calibration procedure is a must where the accurate knowledge of the location of the transmitters and their identifies are acquired and will be stored in a database. However, the traditional manual calibration procedure might be tedious, inefficient, and prone to error. To cope with the problems of manual measurements, some calibration procedures have been proposed using a robot for manual or automated calibration. Nevertheless, controlling a robot may be less convenient and less intuitive. Thus, this thesis proposes a new approach that simplifies the procedure to using only a smartphone, which was equipped with special hardware like a Time-of-Flight camera, offers surveyors more degrees of freedom, and makes the procedure more convenient and less costly. In this thesis, four LEDs modulated in four distinctive frequencies are used as the example VLP system to be calibrated. The outcome of this procedure is a map marked with the precise position and frequency of the LED. To achieve this, the map of the environment is built using RTAB-map, a simultaneous localization and mapping algorithm, and the frequency is measured based on the rolling shutter principle. The algorithm of the mapping of the light can be briefly explained by first find the position of the LED in the coordinate system of the mobile phone, then convert it to the map coordinate system. With this approach, the LED modulation frequency can be recovered with an accuracy of 20Hz and the position estimation error of the LED is around one decimeter. Additionally, the performance compared to the robot calibration approach and the user experience analysis are also discussed. Finally, this thesis also points out some shortcomings of the existing system and suggests corresponding improvements.