基于光谱角时序不变性的红外目标识别

为了识别空间目标与气球诱饵,提出了基于光谱角时序不变性的红外目标识别方法。通过分析证实了空间目标与气球诱饵在温度时变特性上存在差异。为了反映目标测量功率温度时变本质,消除传感器与目标距离以及等效投影面积的时变耦合对测量功率的影响,提出了利用时序光谱角作为光谱向量的时序变化量度。该量度能够刻画光谱曲线的时序变化,反演物体温度时变特性,可以作为温度时变特性不同的目标和诱饵之间聚类识别的有效判据。仿真结果表明：该方法能够有效识别空间目标与气球诱饵;通过融合时序光谱特征,为空间目标识别提供了一种新的机制。

IR target recognition based on invariant sequential spectral angle

An infrared target recognition method based on the invariance of spectral angle in time sequence is proposed to recognize spatial targets and balloon decoys.The difference between spatial targets and balloon decoys on the time-varying characteristics of temperature is gained.In order to reflect the time-varying characteristic essence of target measurement power temperature and to eliminate the time-varying coupling influence of the sensor,target distance and the equivalent projective area on the power measurement,the sequential spectral angle is presented as a sequence varying measure of the spectral vector.The measure can depict the sequence varying of the spectral curve and unti-demonstrate the time-varying characteristics of object temperature by the measurement,and it is regarded as an efficiency criterion of clustering recognition concept between targets and decoys.Simulation result shows that this method could effectively recognize spatial targets and balloon decoys.It provides a new mechanism for space target recognition by merging the sequential spectral characteristics.