一种单像外方位元素的稳健估计方法

获取影像外方位元素是摄影测量学一直要解决的问题,传统的单像空间后方交会利用测量控制点来计算外方位元素,但是很多情况下用于计算的数据是包含粗差的,粗差出现的概率约占1%~10%。本文设计了一种稳健的单像空间后方交会计算方法,把控制点坐标作为未知参数对待并线性化共线方程,最终达到了剔除粗差的目标,提高了计算结果的可靠性,可得到控制点及像点坐标的精度评价。

A robust method to calculate exterior orientation elements of image

How to acquire elements of exterior orientation has been the problem to be solved in photogrammetry.Spatial resection used GCPs(Ground Control Points) to calculate the parameters of exterior orientation,but in many cases the data for calculation contains outliers.Outliers may account for 1% ~ 10% in all.A robust method of detecting the outliers and calculating the image parameters were presented in this paper.It treats the GCPs as the unknown parameters and linearizes the equation,achieves good results and deletes the outliers finally.This method improves the reliability of calculation and can get the precision of GCPs and pixel points coordinates.