A Truly 3D Visible Light Positioning System Using Low Resolution High Speed Camera, LIDAR, and IMU Sensors

Camera based visible light positioning (VLP) can be promising for indoor positioning applications powered by the widespread use of light emitting diodes (LEDs) in lighting applications. Most of the VLP works in the literature have either implemented two-dimensional (2D) indoor positioning or assumed that the height of the ceiling is either fixed or known. In this paper, the real-time height of the ceiling from the receiver is taken into consideration to make the three-dimensional (3D) positioning possible in practical use case scenarios. A low cost light detection and ranging (LIDAR) sensor has been used to measure the height of the ceiling in real time to estimate the location in 3D space using a single LED transmitter. Additionally, though camera based indoor positioning using the rolling shutter effect, offers high accuracy, its reliability at long distances and robustness against the line of sight (LOS) blockage are still questionable. In this paper, a low resolution, high frame rate camera is proposed for an indoor positioning system with centimeter-scale accuracy. Direct detection has been utilized instead of the rolling shutter effect thus making the system more robust and reliable. A gravity virtual sensor has been used to compensate for the error caused by the tilt of the smartphone. The average 3D positioning error in the proposed VLP system is 6.8 cm. The positioning latency of the proposed system is 54 ms. To the best of the Authors' knowledge, this is the first reported truly 3D VLP system that works without any prior knowledge of the ceiling height and irrespective of LED shape, size, and arrangement using a single LED. The proposed system is also robust against any partial light blockage by the user's body and it can also provide coarse location information without any direct LOS between the LED and the camera using reflected lights from the surroundings.