Planar Trajectory Estimation of UHF RFIDs: Performance in Real-World Scenarios

Abstract

This contribution presents phase-based planar trajectory estimation for real-world measurements of movements from tagged objects with passive Ultra High Frequency (UHF) Radio Frequency Identification (RFID) tags. A Motion Capture (MoCap) system serves as a source of high-precision ground truth for validation of the estimator’s performance. Since real-world data contains complex movements in three dimensions, one needs more computationally complex tracking algorithms. We demonstrate in this work that our two-dimensional phase-based particle filter framework is capable of performing planar tracking (x-y plane) of 3D movements by predefining height information for the whole filtering process, which is, in most applications, sufficient Comparison of two trajectories, moving a tagged box on a trolley and in the hands of an operator, shows good results on planar trajectory estimation.