Performance of Target Acquisition Systems with Ordered Commitment

The work presented here provides methods for computing the probability that a cued acquisition system will commit to the correct target based on the performance of its constituent subsystems. A general expression for probability of acquisition (the probability that the target is among the K ≥ 1 best ranking candidates) is derived. Various decision rules for fusion of kinematic- and signature-features are compared. To aid in designing complex systems, approximations using logistic regression are demonstrated.