Incorporation of Uncertainty into Level 2 Fusion Association Metrics
Stephen Craig Stubberud, Kathleen Ann Kramer · 2005
Level 2 data fusion, also referred to as Situational Assessment, can be considered the determination and interpretation of relationships between objects. In a battlespace environment, these related objects can be considered Level 1 target tracks or existing Level 2 objects or groups. The Level 2 state used is comprised of group kinematics, group composition, formation, and region of influence. To develop an automated Level 2 fusion approach, association metrics were defined with mathematical rigor. These association metrics included a group distance measure, a cardinality distance measure and gap metric angle measure for composition, and an area metric that can be used for region of influence. The initial development of these metrics did not include the ability to process target uncertainty information. In this work, the target uncertainty information is incorporated successfully into the aforementioned metrics. The defined uncertainty models are based on the Gaussian covariance for the target kinematics and the use of a Bayesian taxonomy for reporting classification information.