Research On The Method Of Multisensor Target Information And Tracking Fusion

We systematically presented the basic principles, the sense of study, the situation of study of domestics and abroad, and the current study emphases and the direction of development of data fusion. In addition, we also discussed the system model, structure and main algorithms of data fusion. Target tracking play an important role in the military stage. Only if the targets are precisely and reliably tracked, it can be hit. Based on the discussion of several usually used tracking doors, for the study of large scale scan with infrared, we detruded the better performance tracking doors in which we have known data relevancy is the best. In this way , we reduce the echo that does not come from targets in the tracking door, and in the end, we realized the aim of improving the performance of tracking systems. We study the application of neural network in targets tracking. On the basic of the discussion of optimistic data compression based radar and infrared information fusion and flight path fusion, we added the BP neural network and RBF neural network, and simulated. The results of the simulation showed: the method of tracking with multiple sensors can avoid the limitation of single sensor, and improve the precision of targets tracking. At the same time, we can conclude that RBF neural network have better approaching ability than BP neural network and spend less time in training, and can get high precision from the simulation. Last, we study the application of fuzzy in the mobile multitarget tracking. Fuzzy technique can introduce people's experience in the tracking systems, which improve the precision of position estimation. The results of simulation prove that it is efficient.