Distributed Tracking Fidelity-Metric Performance Analysis Using Confusion Matrices.

摘要

Distributed target tracking and identification is an important element of operational environments. In this paper, we develop a fidelity metric of track purity assessment using confusion matrix (CM) fusion. We assess individual distributed tracker track purity associations for a multitarget scenario from two platforms. The fidelity metric for each tracker is combined using the CM fusion for track decision-level analysis to aid in the joint assessment of the track quality. CM fusion enables the estimation of the combined quality of the distributed tracking scenario and can be used for any fidelity metrics based on cardinality. In a distributed multisensor multitarget scenario, we demonstrate the fidelity metric CM fusion for enhanced tracking performance evaluation.