Multisensor tracking and performance assessment is a critical element for the development of any sensor system. In this tutorial, the design trade-offs for multisensor tracking systems will be explored including measurement-based and track-based architectures. In addition, the successful development of any system requires good system-level performance metrics, and performance assessment of multisensor-multitarget tracking systems is particularly challenging. The challenges associated with track-to-truth assignment and potential solutions will be presented along with a myriad of possible metrics that can be computed and analyzed in the development of a multisensor tracking system. Topics for this tutorial include:
- Multisensor tracking architectures
- Track-to-truth assignment techniques
- Multisensor tracking metrics
- Practical applications to radar, sonar, and IR systems
W. Dale Blair is currently a principal research engineer with the Georgia Tech Research Institute (GTRI) and a GTRI Fellow. Since joining GTRI in 1997, Dr. Blair has led a multi-organizational team in the development of multiplatform, multisensor, multitarget benchmarks for both air defense and ballistic missile defense. His research interests include radar signal processing and control, resource allocation for multifunction radars, multisensor resource allocation, tracking maneuvering targets, and multisensor integration and data fusion. Dr. Blair’s research is reported in over 250 publications. He has serves or has served as a consultant for the Applied Physics Laboratory of the Johns Hopkins University, Raytheon Systems Company, Numerica Corporation, Sparta Corporation, Toyon Corporation, and Vectraxx, Inc.
Dr. Blair is co-editor of the Multitarget-Multisensor Tracking: Applications and Advances III. He is internationally recognized for conceptualizing and developing benchmarks for comparison and evaluation of target tracking algorithms. A tracking benchmark is a computer simulation program that includes the salient features of the sensor system of interest and provides a “level playing field” for the evaluation and comparison of tracking algorithms. Dr. Blair led the development of NSWC Tracking Benchmarks I and II and originated ONR/NSWC Tracking Benchmarks III and IV. NSWC Tracking Benchmark II has been used in the United Kingdom, France, Italy, and throughout the United States, and the results of the benchmark have been presented in numerous conference and journal articles. Dr. Blair is also the originator, coordinator, and primary lecturer in the short course Target Tracking in Sensor Systems that is offered annually through the Department of Professional Education and Distance Learning at the Georgia Institute of Technology.
Darin Dunham has conducted extensive research in developing and improving target tracking algorithms. He has worked with several internationally recognized target tracking experts, including Sam Blackman, Oliver Drummond, Peter Willett, Dale Blair, and Mark Levedahl. The list of algorithms on which he has worked is the Multiple Hypothesis Tracker (MHT), Probabilistic Multi-Hypothesis Tracker (PMHT), Best Hypothesis Tracker, and several others. Additionally, he has worked on the many variants of the PMHT. Additionally, Mr. Dunham has done extensive testing and development in the Ballistic Missile Defense (BMD) Benchmark, Integrate Air-Missile Defense (IAMD) Benchmark, Multiple Input, Multiple Output (MIMO) Benchmark, Air Autonomous Vehicle (AAV) Benchmark, and the Joint Composite Tracking Network (JCTN) Benchmark. This work included everything from infrastructure development to beta testing to integrating various tracking algorithms.