Infrared target tracking using multisensor data fusion

Abstract

The infrared (IR) sensor provides the azimuth and elevation angle measurements of the target. For 3-D tracking, at least two IR sensors are needed. The conventional method puts two sensors in good observation position to gain high precision target track. However, the performance of the method is limited because of instability and observable limit of IR sensors. Therefore, more IR sensors are required to improve efficiency of the tracking system. Multisensor data fusion algorithm proposed in this paper is a novel approach to handle measurements from multiple IR sensors. Measurements extracted from every IR sensor by image processing are put into the extended Kalman filter. Then intersection results of measurements from two sensors in acceptable geometrical position are computed. Every intersection result is assigned a weight factor that represents the performance of intersection using fuzzy logic techniques. The fused estimate of the target is obtained by using a weighted average method to all the intersection results. The simulations with Monte Carlo methods show that the proposed algorithm can fuse the target tracks effectively and accurately. Compared with conventional algorithm, the new algorithm can provide higher precision and more robust estimate of the target.