A Three-Stage Hybrid Localization Method Fusing Bluetooth Low Energy and Visible Light Positioning

Bluetooth low energy (BLE) can achieve low-cost positioning with moderate accuracy, while visible light positioning (VLP) can achieve accurate positioning. However, VLP requires each light-emitting diode (LED) to be equipped with communication capability, which can be expensive especially for large area positioning. To address this issue, this paper proposes a three-stage hybrid localization (TSH-Loc) method fusing BLE and VLP. In the considered model, standard LEDs without communication capabilities are installed on the ceiling, and the user is equipped with both Bluetooth and a camera for self-localization. In particular, a visual fingerprint database is first established. Then, a match approach is used to determine the luminaire identification (ID) based on the Bluetooth received signal strength indicator (RSSI) and visual image taken by the camera. Finally, the user’s pose and location are obtained by utilizing the contour of the strip light captured in the image. TSH-Loc achieves high positioning accuracy without relying on specialized LEDs, thereby simplifying system deployment and enhancing overall positioning efficiency. Experimental results demonstrate that the proposed TSH-Loc achieves an average positioning accuracy of 10.8 cm in two dimensions and 11.9 cm in three dimensions, representing an improvement of over 63% compared to the baseline.