相机参数在轨标定是无地面控制点条件下提高定位精度的关键环节, 利用立体影像进行空中三角测量光束法平差, 是实现相机参数在轨标定的有效途径, 但要保证所建立的航线模型无因姿态变化率而造成的系统变形。本文针对“天绘一号”卫星有效载荷的特点, 采用基于LMCCD影像的EFP光束法平差, 按反解空中三角测量原理进行三线阵相机的重组, 实现对相机参数在轨标定。利用LMCCD影像进行相机参数在轨标定后, 天绘一号卫星无地面控制点条件下定位精度上下视差减小至0.6 pixel。定位精度从9 m提高至7.4 m, 其中高程精度从4.7 m提高至2.7 m, 试验结果表明, 所提方法实现了国内相机在轨标定参数。

On-orbit calibration of camera parameters is a key step in the process of enhancing positioning accuracy without the need of ground control points. Triangulation-based bundle adjustment is an efficient method for calibrating the geometric parameters of a camera by incorporating stereo imagery. However, it is necessary that the established strip model should not cause system deformation owing to the rate of attitude variation. In this study, characteristics of the payloads present in the TH-1 satellite were considered, and EFP bundle adjustment based on Line-Matrix Charge-Coupled Device(LMCCD) imagery was applied. Three linear array cameras were restructured by adopting inverse solutions of the aerial triangulation principle, and on-orbit calibration of camera parameters for the TH-1 satellite was thereby realized. The upper and lower parallax of the positioning accuracy is reduced to 0.6 pixels, and the positioning accuracy is increased from 9 m to 7.4 m. The elevation accuracy is increased from 4.7 m to 2.7 m. Obtained results indicate that without incorporating ground control points, the positioning accuracy of the TH-1 satellite imaging system can be significantly enhanced through on-orbit calibration of camera parameters by employing LMCCD imagery.