Recursive tracking algorithms: from the kalman filter to intelligent trackers for cluttered environment
Yaakov Bar‐Shalom · 2005
Tracking in a cluttered environment is characterized by uncertainty in the origin of the measurements. The techniques that are currently in use for association of measurements to measurements (track formation) or measurements to tracks (track maintenance) do not have the built-in capability of assessing their own performance in real time. An intelligent tracker is one that can automatically assess the credibility of its output, i.e., whether it has a target in track, as well as the accuracy of its estimates. A survey of the existing algorithms for data association and tracking is given together with their main features capabilities vs. complexity. A new recursive algorithm, based on the Interacting Multiple Model Probabilistic Data Association Filter, which can start up tracks, maintain them in the presence of target maneuvers and evaluate a "true target probability" for each track is discussed.