基于RPC模型的IRS-P5影像正射校正

IRS-P5 Satellite Image Ortho Corrections Using RPC Model

为了提高有理函数(rational polynomial coefficient,RPC)模型校正模型的定位精度,以IRS-P5为例,通过控制点及补偿模型的添加,比较分析了影像正射校正的精度变化.发现使用1个控制点比不使用控制点,其校正精度从55.80m提高到5.80m;当控制点数达到6个时,其定位精度接近于5.0m,之后影像定位精度便不会再随控制点数的增加而提高;使用1个控制点和补偿模型,可将定位精度提高到4.0m左右.实验表明:可通过添加少量控制点提高RPC模型的正射校正精度,在缺少控制点的情况下,使用补偿模型亦能使RPC模型达到相当的影像校正精度.

Optimizing the RPC（Rational Polynomial Coefficient） model can improve the positioning accuracy of the RPC correction model.Tacking an IRS-P5 image for example,in the cases of using 0-9 control points or using compensation model,we analyze the accuracy change of the image ortho correction.The results show that the accuracy can be improved from 55.80m to 5.80m when we increase the control points from 0 to 1;while the accuracy is close to 5.0m when the number of control points is more than 5;In addition,the accuracy reach 4.0m when we use 1 control point and compensation model.It indicates that the accuracy of ortho correction of RPC modal can substantially increase by adding control points,also can reach high correction accuracy using compensation model in the cases of missing control point.