针对具有推扫机制的狭缝体制成像光谱仪受卫星平台俯仰、 侧滚、 偏航等复杂运动的影响, 导致光谱数据精度降低的问题, 在成像光谱仪运动成像光谱微分动态成像退化仿真方法的基础上, 提出基于点扩散矩阵的光谱退化理论。按照八邻域掺杂模型, 考虑卫星平台运动的时变性, 不同目标像元的点扩散矩阵不同, 符合伪互相关运算的概念, 因此提出运动成像的伪互相关光谱退化理论。其中, 点扩散矩阵由星上POS数据运用微分像移理论计算八邻域掺杂像元的平均掺杂比得到, 在用模拟POS数据曲线仿真计算点扩散矩阵时, 发现依据掺杂影响的大小, 点扩散矩阵可以简化以便减小运算量。明确阐述光谱运动成像的伪互相关退化理论的表述和计算方法, 对退化仿真和计算的结果从图像维和光谱维分别进行了定性和定量的效果评价, 并使用结构相似度参数展示了退化图像与原始图像的相似性。运动成像的伪互相关光谱退化理论完善了已有的基于卫星平台复杂运动的光谱数据退化问题, 仿真结果表明此种退化理论完全适用于解决卫星平台复杂运动条件下的光谱数据退化问题。

The present paper proposed a spectral degradation theory based on the point spread matrix, which is developed from the simulation algorithm of degradation based on differential dynamic spectral imaging for moving imaging. According to the eight neighborhood mixing model, considering the varying movement with time of the satellite platform, the situation that different object pixel has different point spread matrix is corresponding to the concept of the pseudo cross-correlation theory. Thus, the pseudo cross-correlation degradation theory of moving spectral imaging is presented. In this method, the point spread matrix is constructed by the mean mixing ratio of eight neighborhoods, which can be calculated by the POS data on satellite using the differential imaging theory based on image motion matrix. When calculating the point spread matrix with the simulating POS data curve, we found out that the point spread matrix can be simplified according to the importance of mixing effect to reduce calculation. Then, the presentation and computing method of the pseudo cross-correlation degradation theory of moving spectral imaging is clearly articulated, and the result of degradation simulating and calculating is evaluated qualitatively and quantitatively from imaging dimension and spectral dimension respectively, and the similarity between the degradation image and the initial image is revealed using the SSIM parameter. The pseudo cross-correlation degradation theory of moving spectral imaging improves the existing problem of spectral imaging degradation based on the complex movement of satellite platform. The simulation results show that this kind of degradation theory is suitable for solving the problem of spectral imaging degradation based on the complex satellite platform movement completely.