基于G-N算法的解析空间坐标旋转变换

On the Coordinate rotation Transfer Based on G-N Algorithm

摄影测量中,用像点坐标解求相应地面点坐标时,它涉及到坐标系统之间的坐标变换,这种变换是非线性的,理论计算时,一般是在假设旋转角为小角度的基础上进行的.本文提出一种适用于任何旋转角度的非线性最小二乘空间三维直角坐标转换算法,通过解多元函数极小值求取七参数最佳估值.与目前常用的一些方法相比,具有计算简便、收敛速度快、不依赖于七参数初值、便于程序实现等优点,并通过模拟算例验证了该方法的可行性.

In photogrammetric surveying, the algorithmic finding for coordination on earth with picpointed coordination involves nonlinear transfer of coordination among coordinate systems. Theoretical arithmetic is generally carried out under the supposition that the angle of rotation is at minimum. This paper put forward a nonlinear algorithm of coordinate rotation transfer, which is of least squares and PLOT 3D. Compared with other general methods at present, the algorithmic search for the best estimated value of seven-parameter through the minimum multivariate function is more convenient, efficient and less dependent on the default seven-parameter. It also helps the system function properly. The paper has testified its feasibility through an algorithmic case.