Research On Combination Adjustment Of Strict Geometric Model And RFM Model Of Optical Satellite Remote Sensing Images

With the continuous development of satellite remote sensing ground observation technology,optical satellite remote sensing images have become an important data source in various fields of social economic construction and national defense construction,and their geometric processing accuracy requirements are getting higher and higher.This article takes optical satellite remote sensing images as the research object,and starts with its strict geometric model and general geometric model.Based on the improvement of the two models,a strict geometric model and a rational function model(RFM)are proposed.The combined adjustment scheme and its implementation method provide a new technical approach for the precise geometric processing of optical satellite remote sensing images.The main work and innovation of the paper are as follows:(1)Realize the strict geometric model adjustment method of optical satellite remote sensing images and carry out simulation experiments to obtain a quantitative relationship between the accuracy of the strict geometric model adjustment and the initial value accuracy of the external orientation elements of the image.In the ideal situation where the image is not distorted,strict geometric model adjustment can well eliminate the influence of the error of the external orientation elements of the image,but the accuracy of the adjustment is still affected by the accuracy of the initial value of the external orientation elements.When the image contains distortion,the error in positioning after strict geometric model adjustment increases linearly with distortion.(2)Implement the conventional RFM adjustment method with affine transformation terms and perform simulation analysis on its performance.The results show that the positioning accuracy of the method depends on the accuracy of the strict geometric model that generates the initial RPC parameters.As the error of the external azimuth element in the strict geometric model increases,the positioning accuracy of the conventional RFM adjustment method also decreases significantly.For complex distortions present in optical satellite remote sensing images,conventional adjustment methods has poor effect.(3)In view of the shortcomings of conventional RFM adjustment methods,this paper proposes an RFM adjustment method with Fourier polynomial compensation terms based on the approximation characteristics of Fourier series,which can theoretically accurately fit any form of distortion.The fitting test shows that the simulated distortion generated by the Fourier polynomials on the Brown model and the Ebner model can reach the fitting accuracyof one thousandth of the pixels.The two RFM adjustment methods were used to compare the fast-track stereo image data of the Quick Bird.Compared with the conventional RFM adjustment results,the plane positioning accuracy after RFM adjustment with Fourier compensation is improved by 11.6%.(4)In view of the respective deficiencies of a single model,this paper proposes a combined adjustment scheme of strict geometric model and RFM model and its implementation algorithm flow.The experimental results of simulation data and real data show that the combined adjustment can not only significantly reduce the accuracy requirements of the observation values of the external azimuth elements,but also because of the excellent distortion compensation performance of the Fourier polynomial,the iterative adjustment calculation of the strict geometric model and RFM model can at the same time achieve high-precision image orientation and positioning.Experiments on the SPOT5out-of-track stereo image data show that the plane and elevation positioning accuracy after combined adjustment can reach a level better than the ground resolution of one pixel.