Unified Robust Multisource-Multitarget Targeting and Identification

Period of Performance: 04/12/2000 - 04/12/2002

$750K

Phase 2 SBIR

Recipient Firm

Scientific Systems Company, Inc.
500 West Cummings Park Array
Woburn, MA 01801
Principal Investigator

Research Topics

Abstract

Scientific Systems Co., Inc. (SSCI) and its subcontractor Lockheed Martin Tactical Defense Systems (LMTDS) propose a direct, systematic, and principled generalization of Bayesian filtering and estimation theory to the problem of multisource, mutitarget, multi-evidence unified joint detection, tracking, and target I.D. for possilby low-observable ground and air targets. Recursive Bayesian nonlinear filtering and estimation is the mosts accepted standard for developing algorithms that are optimal, theoretically defensible, and practical. However, the sources of any optimal-Bayes technique hinges on a number of subtle issues. If any of these issues are not properly addressed, optimal-Bayes filters will often be non-robust. In recent years, shortcomings in conventional Bayesian thinking-and corresponding desires to expand the conceptual scope of Bayesian methods to deal with them-have become evident. The SSCI-LMTDS approach is capable of addressing: (1) the ambiguity that is present in real-word ATR problems involving sources such as SAR, HRR, and datalink messages; (2) the real-time computational issues in single-target filtering, especially in low-SNR regimes; (3) the unique conceptual and computational difficulties associated with optimal multitarget filtering. This approach finds its basis in finite set statistics (FISST) a methodology developed at LMTDS. Specific Phase II tasks are: (i) continuation of a single-target tracking and I.D. using radar and HRRR data from the ARTI database. (ii) develop multisource, multi-evidence unified tracking and identification algorithms. (iii) develop a computationally tractable Multitarget Tracking and I.D. scheme. (iv) Final report, software delivery and Phase III recommendations. The project team includes Dr. Ronald Mahler of Lockheed Martin TDS the originator of FISST and Veridian. Lockheed Martin will provide technical and commercialization support in the application of these technologies. Veridian will provide assistnace in data procurement and interpretation for