Feature-Aided Tracking (FAT)

Cooperative engagement and cooperative sensing tasks require the development of accurate and persistent tracks. Though much work to date has focused on tracking using kinematic information only, a wide variety of feature information on the targets may be available to the tracker. Combining this feature information with kinematic information can improve track accuracy and persistence, especially in challenging environments. Toyon Research Corporation proposes a dual-layer solution for feature-aided tracking to deal with both short-term individual measurement associations and longer-term track-to-target associations. A Multiple Hypothesis Tracker (MHT) will be used in conjunction with a Bayesian network to model feature information and possible inferences garnered from this information in a way that promotes improved measurement-to-track association. Toyon’s Tracked Object Manager (TOM) will handle feature database management and use its track stitching algorithms to maintain long-term continuous track, even in situations where features do not singly provide much discrimination between targets. These layers may be implemented in a distributed and decentralized environment. Toyon will also design a test scenario in order to test these algorithms.