航空倾斜摄影系统多相机姿态安装误差检校

针对航空倾斜摄影系统多相机间相对关系的姿态误差,提出一种利用下视影像密集匹配点云约束的联合平差多相机姿态安装误差检校方法。文中方法以共线方程为基础,通过联立倾斜立体像对连接点、下视影像密集匹配点云数据提供的高程约束虚拟观测值,采用非线性最小二乘SQPM算法求解倾斜影像的外方位元素改正数,从而解算下视相机与倾斜相机间的姿态安装误差。为了保证连接点可靠性,提高虚拟观测值的有效性,采用下视影像密集匹配的点云约束倾斜影像立体连接点匹配和分布优化过程。利用SWDC-5倾斜摄影系统获取的数据进行实验,通过原始姿态检校参数将下视影像生成的密集点云投影至倾斜立体影像上,在像方有22个像素左右的误差,采用文中方法检校后,像方误差减少到0.29个像素。

Calibration method for installation attitude error between multi-cameras of airborne oblique photogrammetry system

Aiming at the attitude parameter error between multi‐cameras of airborne oblique photogrammetry system ,this paper proposes a method based on combined adjustment which uses the dense matching point cloud from nadir image for a constraint .This method is based on collinear equation , by integrating tie points ,virtual observation from point cloud data derived from nadir image based on dense matching as an elevation constraint , and using the principle of least‐squares adjustment with SQPM algorithm to solve the corrections of exterior orientation elements of oblique images .Then high‐precision attitude parameter can be obtained .To ensure the reliability of the tie point ,and the effectiveness of virtual observations ,the point cloud derived from dense matching is used for constraining tie point matching and distribution optimization processing . Images from SWDC‐5 oblique photography system are used for experiments .By projecting the point cloud data from nadir image and using dense matching to stereo oblique images wih the initial attitude parameter ,there are about 22 pixels error in image space .After calibration based on the proposed method ,the error is reduced to 0.29 pixels .