Measurement-to-track association using fuzzy logic compared to the chi-squared metric

The concept of target tracking, a part of Level 1 data fusion, is to combine measures from various sensors to form a coherent picture of the scene. One of the major concerns with this problem is the association of the various measurements with the existing target tracks. The standard approach to the kinematic association problem is the chi-squared metric, a weighted inner product of the residual formed by an estimated measurement and the true measurement. The chi-squared metric is a proven tool that requires both the measurement and track to be defined by Gaussian probability density functions. At times the Gaussian is not always the best approximation. In this paper, we demonstrate that a developed fuzzy logic association technique has the capability to emulate the chi-squared metric. Since fuzzy logic does not require the Gaussian assumption, we can utilize the approach when the Gaussian assumption fails.