Non-Point Visible Light Transmitter Localization based on Monocular Camera

As German proposed in 2013, the 4.0 industrial revolution would influence the world with massive devices and actuators in multiple applications and scenarios. According to Cisco forecasting, by 2030, the predicted number of the connected devices on the Internet will reach up to 500 million. Under this situation, coupled with the effect of Beyond 5 Generation (B5G) and the expected features of 6 Generation (6G), there is the growing demand for massive Device-to-Device (D2D) communication on the Internet of Things (IoT), thereby, the standards of the communication performances are getting stricter and they may depend on the performance of localization under certain circumstances, especially indoor localization. Considering the cost of infrastructure, localization precision and some non-ignorable factors such as the information security and abundant spectrum resources, visible light positioning (VLP) is adopted to accomplish the better indoor localization. Among the existing localization technologies including Time of Arrival (TOA), Difference Time of Arrival (DTOA), Angle of Arrival (AOA), Received Signal Strength (RSS), camera and fingerprint, camera localization is a lower cost and higher speed method with the simper operation[1]. Particularly, monocular camera is highly appreciated because of the low cost and the simple calibration[2-4]. But it has to handle the massive data and the corresponding updates, which leads to more complex computation comparing to binocular camera localization with simpler geometric localization principle[5]. Therefore, the advantages of two methods are combined in this paper. Meanwhile, although the existing transmitter with kinds of shapes, there  are still lots of researches only focusing on the point-transmitter, causing the low practicability of the localization method. Therefore, non-point transmitter will be considered and utilized, which only one light is needed based on the monocular camera.