机载激光雷达点云与定位定姿系统数据辅助的航空影像自动匹配方法

本文针对LiDAR点云和POS数据辅助航空影像的连接点自动匹配，设计了用于相关系数匹配的最佳匹配点位快速搜索算法；提出一种基于虚拟地面控制点的相机安置角误差自动检校方法，在此基础上自适应确定影像匹配搜索范围。试验结果表明，本文提出的最佳匹配点位快速搜索算法可在保证匹配正确性的情况下节省约25%的匹配耗时；相机安置角误差补偿方法能够有效地提高由POS数据计算的影像外方位元素的精度，从而明显改善同名点匹配时的点位预测精度；本文算法能处理多分辨率、多视角的交叉飞行航空影像，获得高精度的匹配结果。

Automatic Tie Point Matching of Aerial Images Acquired by LiDAR System

Aerial images tie points matching is conducted with the assistance of LiDAR points and POS data. Fast searching algorithm for obtaining conjugate point in correlation matching is devised. An automatic boresight misalignment calibration method based on virtual ground control points is adopted, on base of which the searching range of image matching is adaptively determined. The experimental results indicate that the fast searching algorithm proposed in this paper can reduce 20% of the matching time while ensuring the matching correctness. The camera boresight misalignment calibration method can effectively refine the accuracy of exterior orientation elements of images derived from POS data, and significantly improve the precision of predicted point position for tie points matching. The matching algorithm proposed in this paper can process multi-scale, multi-angle, cross-flight aerial images and obtain a high matching accuracy.