弹道目标检测与识别问题是天基红外预警系统的核心难题之一.针对主动段弹道目标的检测和识别问题,分析了传统的基于空间和时间特征信息的弹道目标检测与识别方法.利用目标辐射空间与光谱的一致性,提出一种基于成像光谱技术的主动段弹道目标检测与识别方法,将空域目标检测和谱域目标识别两个环节进行联合处理.实验证明,该方法应用于复杂背景下低信噪比的红外弱小目标图像序列能得到较理想的结果,算法检测概率高、虚警概率低、具有较强的实时性.

Ballistic target detection and recognition is one of the core difficult problems of Space-Based Infrared System (SBIRS).Aiming at the problem of detection and recognition of boost-phase ballistic target,the traditional methods based on spatial and temporal characteristic information were analyzed,and the disadvantages of the methods were pointed out.According to the consistency of the target radiation in spatial dimension and spectral dimension,an integrated method for ballistic target detection and recognition was put forward based on imaging spectral technology,in which integrated treatment was made for target detection and recognition in spectral domain spectral domain.Experiment results show that this method can obtain good result toward low SNR infrared image sequence under complex background,the algorithm has high detection probability,low false-alarm probability and fine real-time performance.