基于嫦娥四号月球车图像的地形遮挡高精度预报试验研究

Experimental research of precise terrain occlusion prediction algorithm based on image sequences of the Chang’E-4 lunar rover

为进行天际线遮挡角的准确预报,提出了月球车图像的地形遮挡角高精度计算方法.首先阐述了月球车图像的地形遮挡角预报流程,然后重点分析了月球车相机在轨几何标定、基于闭合区域网平差的月球车相机在轨位姿参数估计、月球车位姿参数高精度计算和天际线自动提取等关键技术.最后,通过我国嫦娥四号月球车的专项试验,验证了月球车图像的地形遮挡角高精度计算方法具有较高的地形遮挡角预测精度和应用稳定性.

An accurate terrain occlusion prediction algorithm based on image sequences obtained by the Chang’E(CE-4) lunar rover stereo cameras is proposed. First, the proposed algorithm is briefly introduced. Next, the key procedures, including on-orbit camera calibration, high-precision camera pose estimation based on a loop-closed bundle block adjustment model, high-precision rover pose estimation, and automatic extraction of the skyline, are explained. Lastly, the conclusions are drawn based on the experimental results of China’s Chang’E-4 lunar rover. The proposed algorithm can achieve high accuracy and application stability in the occlusion prediction of the skyline.