论文部分内容阅读
天基红外预警卫星远距离探测时获取的弹道目标特征信息匮乏,代表物质固有属性差异的光谱信息可作为目标识别的主要依据。将尾焰特征光谱信息作为识别的重要手段,综合考虑尾焰光谱吸收特性和特征光谱提取原则,采用改进的向前和向后间隔偏最小二乘法建立特征波段提取模型,以新型自适应变权重光谱相似性测度(SAVM)实现目标与特征光谱数据库的匹配,提出了基于尾焰特征光谱的主动段弹道目标识别方法。仿真实验进行了特征波段提取与SAVM优越性的验证,相较于全波段光谱匹配识别法,提出的方法所需数据量更小、识别精度更高。研究内容可为红外预警卫星系统优化探测识别能力提供有意义的参考。
The information of ballistic target acquired during long-range space-based detection of space-borne infrared early-warning satellites is scarce, and the spectral information representing the difference of the inherent properties of matter can be used as the main basis for target recognition. Taking the characteristic information of the tail flame as an important means of identification, taking into account the characteristics of the end flame spectral absorption and the principle of characteristic spectral extraction, an improved forward-backward and backward-interval partial least squares method was used to establish the characteristic band extraction model. With the new adaptive variable weight Spectral similarity measure (SAVM) is used to match the target with the characteristic spectral database. Aiming at this problem, an active segment trajectory target recognition method based on the tail flame characteristic spectrum is proposed. Compared with the full-band spectral matching recognition method, the proposed method requires less data and has higher recognition accuracy. The research content can provide a meaningful reference for the optimization of detection and recognition ability of infrared early warning satellite system.