利用有理函数模型的几何定位仿射变换方法

A affine transformation method of geometric positioning based on RFM

针对高分辨率遥感影像在应用严密成像模型处理时存在求解参数众多、数值解算精度不高的问题,该文提出了基于有理函数模型(RFM)对几何定位进行仿射变换的方法。首先采用高分辨率遥感影像基于有理函数模型系数进行无控制点直接定位;然后利用控制点变化趋势制定控制点布设方案,并选取最优控制方案进行仿射变换;最后基于像方仿射变换进行误差补偿,从而实现减小高分辨率遥感影像存在的系统误差。以IKONOS遥感数据进行几何定位实验,结果表明:本文方法可以有效地提高高分辨率遥感影像的几何定位精度。

In view of the numerous solution parameters and low precision of the application of rigorous sensor model on the processing of high-resolution remote sensing image,a affine transformation method of geometric positioning based on RFM was proposed.First,the direct location of high-resolution remote sensing image without control point was carried out based on RFM;then the layout scheme of control point was made by the change trend of control points,and the optimal layout scheme was selected for affine transformation;at last,error compensation was carried out based on image affine transformation in order to decrease systematic error.The results of geometric positioning of IKONOS data showed that this method could effectively improve the geometric positioning accuracy of high-resolution remote sensing image.