Application of Smoothing Techniques for Tracking Maneuvering Targets. Multiple Target Tracking in Clutter: New Approaches
Final rept. Jun 1886-May 1992
PENNSYLVANIA STATE UNIV UNIVERSITY PARK DEPT OF ELECTRICAL AND COMPUTER ENGINEERING
Pagination or Media Count:
The project was concerned with development of fast algorithms for clustering and data association and then applying fixed-lag smoothers for tracking maneuvering and non-maneuvering multitargets in clutter. The data association was done by a depth-first search DFS technique which extends naturally to three dimensional validation matrices which occur in Markov models of system parameters for maneuvering targets and also in multiscan correlation. A faster algorithm is then suggested for computing approximately the a posteriori probabilities without generating the data association hypotheses. This algorithm is suitable for implementation with multiprocessors in mobile airborne or seaborne distributed tracking facilities, while the DFS-based approach is suitable for ground-based trackers. Simulation studies have been conducted to demonstrate the advantages of the algorithms developed in this project over comparable ones existing in the literature.
- Computer Programming and Software
- Target Direction, Range and Position Finding