الفهرس | Only 14 pages are availabe for public view |
Abstract Multi-target tracking (MIT) problem is an essential requirement for surveillance systems employing one or more sensors, together with computer system, to interpret an en’ ironment that includes both true target and false alarms. With the recent proliferation and the increasing sophistication of surveillance systems, designers of radar and other systems are recognizing the importance of MIT. This thesis presents a survey of the existing theory and methods for MIT, as well as algorithms for gating and prediction for real-time applications. The basic elements constituting a MIT system are also discussed. Filtering, prediction and estimation are explained. The theories of filtering using fixed a-~ and time-varying coefficients (Kalman) filters are presented. The correspondence between them is also explained. A survey of gating and data association is illustrated. Various kinds of gating procedure are discussed. Some adaptive target tracking algorithms are explained. An adaptive target tracking algorithm with simplest case of adaptive a.B filter is proposed. It composed of three main major functions: track initiation, plot-to-track correlation, and track update. The proposed algorithm discusses these three functions and detailed analysis is given. Track update is presented using the necessary relations for pre-update extrapolation, gain selection, update position and velocity filtering or smoothing, covariance matrix, and post-update are discussed. Implementation and results of multi-target tracking (MIT) USIng simulated radar system input are presented. Finally, some simulated target trajectories data are considered and program output, after processing their data, is given. The advantages of the proposed algorithm are as follows it is a recursive algorithm; simple and easy to implement; it can initiate tracks and maintain tracks in the presence of maneuvers and can also terminate tracks if warranted; it provides optimum estimates of the aircraft’s position and velocity; and it has an excellent multitarget tracking performance without significant increasing in the complexity and formulas. |