点光源反射镜法向标校技术

针对基于反射式点光源进行在轨辐射定标过程中反射镜法向标校建模不够完善的问题,提出基于反射镜与相机几何模型的反射镜法向标校及矢量控制算法.通过解算模型求解相机与反射镜间的几何误差,建立了太阳图像质心坐标与反射镜法向之间的关系,可实现多点自动化标校反射镜法向,提高镜法向标校及系统指向精度.实验结果表明,利用解算后的几何模型反解不同时刻质心坐标进行多点反射镜法向标校,相机观测太阳像素角分辨率标准误差分别为:X轴方向0.02165°、Y轴方向0.01982°,综合角分辨率误差为0.02936°,优于太阳观测器对反射镜法向标校精度.实现了相机观测太阳取代人工借助太阳观测器观测太阳的自动化镜法向标校,扩展了标校灵活度,系统综合指向精度优于0.1°,为固定实验场联网自动化集中控制不同能级梯度的点光源阵列在轨辐射定标和调制传递函数检测奠定基础.

Specular Normal Calibration Technology of Point Source

In the process of on-orbit radiation calibration,based on reflective point light sources,the modeling of specular normal calibration is inadequate.Aiming at this problem,a specular normal calibration and vector control algorithm based on reflector and camera geometry model is presented.The geometrical placement error between the camera and the reflector is solved by solving the model.The relationship between the centroid coordinates of the solar image and the normal direction of the reflector is established.The multi-point automatic calibration of the specular normal is realized.The method improves the pointing accuracy of the system and specular normal calibration accuracy.The experiment results show that the geometric model anti-solves the centroid coordinates at different times,which for multi-point specular normal calibration.The standard errors of the angular resolution of the sun pixels observed by the camera are 0.02165°in the X-axis direction and 0.01982°in the Y-axis direction.The error of the synthetic angle resolution is 0.02936°.The accuracy of calibration is better than that of the solar observer to the specular normal calibration.The method realizes the automatic calibration process of the camera observing the solar image instead of manually observing,and expands calibration flexibility.Comprehensive pointing accuracy of system is better than 0.1°.It lays the foundation for on-orbit radiation calibration and modulation transfer function detection of point light source array of networking automation of fixed experimental sites centrally control different energy levels gradients.