民用航空区域危险目标自动识别方法研究

Research on Automatic Identification of Dangerous Target in Civil Aviation Area

针对传统危险目标识别方法未对危险目标证据进行融合,导致识别结果准确性差的问题,提出基于D-S理论的民用航空区域危险目标自动识别方法。采用ISAR与傅里叶变换实现民用航空区域目标瞬时成像。基于成像结果,引入D-S理论实现危险目标自动识别。通过Harris提取目标特征,利用Mean-Shift对目标进行跟踪,根据D-S理论识别跟踪结果,引入民用航空区域目标危险程度概率赋值函数得到目标危险的概率,通过D-S组合规则合成目标危险证据,基于危险目标证据的融合实现危险目标自动识别。通过实验对所提方法进行验证,实验结果表明,该方法识别精度高,具有可行性。对于民用航空安全管理提出的建议为:掌握整体形势,创新安全管理的相关理念及构建较为完善的民用航空安全管理体系两方面,力求为民航发展提供理论支撑。

Aiming at the problem that traditional dangerous target recognition methods do not fuse the evidence of dangerous target,which results in poor accuracy of recognition results,an automatic identification method of dangerous target in civil aviation area based on D-S theory is proposed.ISAR and Fourier transform are used to realize instantaneous imaging of targets in civil aviation area.Based on the imaging results,the D-S theory is introduced to realize the automatic recognition of dangerous targets.Harris extracts target features,uses Mean-Shift to track target,identifies tracking results based on D-S theory,introduces probability assignment function of target risk degree in civil aviation area to get target risk probability,synthesizes target risk evidence by D-S combination rule,and realizes dangerous target based on fusion of dangerous target evidence.Automatic recognition.The experimental results show that the proposed method has high recognition accuracy and feasibility.Suggestions for civil aviation safety management are as follows:grasping the overall situation,innovating the relevant concepts of safety management and building a relatively perfect civil aviation safety management system,in order to provide theoretical support for the development of civil aviation.