The Target Tracking Technology Research Of Multi Sensor Information Fusion System

In this paper, the technology of target tracking in information fusion system are studied. According to the function division of system and the actual situation of the project, the paper studied and formulated the interface data format between simulation and target tracking, target tracking and target identification, proposed the coordinate transformation, time alignment, track correlation, data refinement methods and strategies of track correlation processing.In order to solve the false association in the nearest neighbor method as track crossing, the paper introduced the method of compare the tracking with the velocity vector direction of the measuring point. Based on the statistical distance of the position, the magnitude of the velocity, the angle difference between measured velocity vector and track speed vector. The associated condition is more reasonable and perfect, and the error of association is improved.In order to meet the requirements of engineering applications, reduce the amount of computation, improve the timeliness of the association, resolve the large amount of calculation, more time-consuming problem of the joint probabilistic data association algorithm, this paper discusses a simple joint probabilistic data association algorithm, which introduces the concept of simple assemble and reduces the sensor number and target measurement of the assemble. So it effectively reduces the number of association events and enhances the timeliness.In order to verify the effect of target tracking, according to the requirement of the project for tracking, this paper put forward a test system, and the test result are analyzed by detailed statistics. The accuracy of target tracking with the integration of processing time, the number of sensors, and complex relationship of wave density factors also be summarized.