一种基于可信度矩阵的多传感器目标识别方法

A METHOD OF MULTI-SENSOR TARGET RECOGNITION BASED ON RELIABILITY MATRIX

为了解决多传感器目标识别中不同等级信息源数据的融合问题,在研究DSmT证据理论的基础上,引入证据可信度矩阵。依据可信度矩阵对证据进行转化,使之可以用传统的方法进行证据融合。将这种方法应用到等级不同的多传感器目标识别中,可以解决传统证据理论只能进行相同等级传感器目标识别的难题。仿真实验表明,该方法提高了目标识别的准确性和有效性。

For solving the problem of fusing information data resources of different grade in multi-sensor target recognition,the concept of evidence reliability matrix is introduced based on DSmT evidence theory.Evidences are transformed according to the reliability matrix and the tradition method is then able to be used for evidence fusing.Applying the approach to the recognition of multi-sensor targets in different grade,the poser in traditional evidence theory that only the sensor target in same grade can be recognised is solved.Simulation experiment indicates that the precision and effectiveness of target recognition are raised with this method.