Abstract:

In tnis paper, a tracking initiation problem of multiple maneuvering targets in a dense multi-return environment is investigated.First, the definitions and properties of main tracking subspace and margin tracking subspace are studied.Then, an algorithm of all-neighbors. Bayes tracking initiation suitd to higher dense multi-return environment is presented.Monte Carlo simulation shows that the algorithm overcomes the shortcoming of modified probabilistic data association filter without mechanism of tracking initiation, improves the ability of tracker differentiating between targets and false alarms greatly, and is an eflective tracking initiation method of multiple maneuvering targets in higher dense multi-return environment.

Key words: multiple maneuvering target tracking, tracking in tia-tion, Bayes estimation