Probabilistic cleaning over trajectory events of mobile RFID objects

Abstract

Cleaning up the trajectory events of mobile RFID objects should take the uncertainty of location and unreliability of event detection into account at the same time. In the paper, we first discuss the rules to distinguish false detection events in RFID object trajectories. Then, as a unified cleaning framework, we establish a probabilistic region connection graph to represent region detection features, region connection relationships, and region transition probabilities between neighboring physical regions. Focusing on interpolating the missing events, we suggest two path-based probabilistic cleaning strategies, namely, the Most Likely Path (MLP) strategy and the Highest Weighting Probability Path (HWPP) strategy.