Research On The Positioning Precision Control Method Of High Resolution Remote Sensing Satellite Image

Research on the positioning precision control method of high resolution remote sensing satellite image can promote the development of space remote sensing and meet the multi-level,diversification needs of the users.With the analysis of imaging model of remote sensing,the optimization and degradation of the positioning precision are studied.The contributions of the thesis are given as follows.1.The imaging models of remote sensing at home and abroad are studied completely.Then the methods to calculate Rational Polynomial Coefficients(RPC)and ground position are highlighted.It provides a basis for further study of the optimization and degradation of the positioning precision.2.The optimization model based on single scene and stereoscopic model are studied.With the number and distribution of ground control points as experimental conditions,using the ZY-3 satellite images as experimental data,this paper analyses merits and demerits of all kinds of optimization models in different conditions and puts forward the selection scheme.3.A self-calibration bundle block adjustment method which can optimize the positioning precision is proposed.With the optimization models of single scene positioning precision,the optimized parameters of the remote sensing images and the ground three-dimensional coordinates of tie points are calculated.The experimental result shows that the method can realize the optimization process and improve the positioning precision.4.Two methods about the degradation of positioning precision are proposed.First,the degradation method based on the virtual grid of points is proposed by using the virtual grid of points as mediums and the size of randomly error in lattice points to adjust the precision.The experimental result shows that it can realize the stability control in a certain scale.Second,the degradation method based on the normalizing parameters of RFM is proposed by adjusting the correction of the normalization offset parameters.Experiments show that the method performs more stability,controllability and practicality.