电视与红外机动目标数据融合的研究

The Study on Data Fusion of Infrared and Video Maneuvering Targets

从跟踪性能、计算的复杂性以及应用范围等方面对红外与电视传感器的三种融合模型进行了比较,得出了来自同一平台上的红外与电视传感器的数据,在各传感器没有自己的数字处理机的情况下,点迹融合模型较其余两种模型实用的结论,在点迹融合模型的基础上,用极大似然融合算法对点迹进行融合处理,在理论上证明了该算法用于此种模型能提高目标状态信息的估计精度,降低测量误差。并给出了本算法用于此种模型在计算机上的仿真结果,从实际应用上进一步说明了本算法的有效性。最后利用卡尔曼滤波实现了红外与电视机动目标的跟踪。

Three fusion models of the infrared and video target are compared in tracking performance, the complexity of computation and the field of the application. Our conclusion is that if the data come from the same platform, the measurement fusion model is more practical than the others. The data are fused by the maximum likelihood algorithm based on measurement fusion model. Using this algorithm and the measurement fusion model, theoretically, the precision of the status of the target can be improved, and the error of the measurement can be decreased. And the result of simulation is presented, which proves the availability of the algorithm in practice. At last, the paper introduces the applications of Kalman filtering and tracking algorithm for infrared and video maneuvering targets.