An Adaptive Filter for Multi-sensor Track Fusion

An adaptive filter is developed for use in multi-sensor surveillance systems for target tracking. The hierarchical architecture consists of local processors and global processor for distributed fusion. A linear Kalman filter is employed in each local processor to track the same target which is described in the reference Cartesian coordinate system with the radar measuring range, bearing and elevation angle in the spherical coordinate system. The global processor utilizes Information Matrix Filter (IMF) with two different process noise levels to combine local processor outputs. A decision logic approach incorporates with dual-band IMF to develop switching capability as adaptive manner to respond the target dynamics. The high-level band IMF is selected for maneuver period and the proper low-level band IMF is chosen for quiescent period. The resulting filter has better tracking performance than each individual IMF. Simulation results are included to demonstrate the effectiveness of proposed algorithm.