Innovative Indoor Positioning System Based on Infrared Light Technology and SVR Model

This work presents an indoor positioning system based on infrared (IR) cameras. The system gathers identity (ID) and 3D position information from an IR tag placed on the tracking target. Compared to other positioning methods, the proposed system offers precise positioning capability without high costs or privacy concerns. The system utilizes a two-stage Support Vector Regression (SVR) model to estimate the height of the IR tag in the real world and calculate the coordinates for positioning. Experimental results indicate that the proposed two-stage SVR model can effectively track the target's movement with an average 3D positioning error of 10.1 cm and achieve 100% accuracy in ID recognition.