Fuzzy logic track classifier

Abstract

In this paper the structure of the track classifier based on the fussy logic is presented. The track classifier is a decision system, which allows determining whether the sequence of radar detections forming a track comes from the real target, or from other objects generating clutter. Human operator easily distinguishes the real tracks from the false ones; however the automaton of this process is not trivial. There are no unique criteria for distinguishing the real and false tracks. In general, the sequence of detections coming from the real track is more regular, that means there are not rapid changes of parameters of motion. The term "rapid change" is not obvious. To determine whether the parameters of motion change rapidly, the fuzzy logic is applied. From this comes the term "fuzzy logic track classifier". The results of the investigation are very encouraging. Applying this method, the number of the false tracks has been reduced more than tree times.