Technical Research On The Improvement Of Positioning Accuracy Of Hingh Resolution Remote Sensing Satellite Image System

With the development of high resolution remote sensing satellite image,demand for higher positioning accuracy of remote sensing image also rises.Positioning accuracy of systematic products mostly depends on the ephemerisdata, attitude data, and star time accuracy as well as the geometrical performanceof remote sensor.There are mainly two types of models for ground processingsystem, namely strict physic model and general experience model. The formerrequires the orbit and attitude parameters of sensors during operation process,however due to the parameter information represents certain physicalsignificance, thus it is confidential and not available. Only RPC(RationalPolynomial Coefficients) model is applicable as a representative of generalexperience model, which will acquire higher accuracy similar with the imageaccuracy from strict model correction with few control points. Using RPC modelto increase the high resolution system positioning accuracy is the focus of thepaper.Firstly, the paper describes the satellite development and research status athome and abroad, analyzes the system errors during satellite imaging, raises thesystem error correction solutions and introduces the RPC model and its accuratesolution. With the aim of increasing the image direct ground positioning accuracy,a new solution to eliminate system errors need to be found and a newcompensation model of system errors to be set up. It has been proved that groundcontrol data is the most effective way to correct system errors by fully using thehigh accuracy control data to get the partial error correction parametersaccurately and establishing correction parameter predict model to estimate theparameters thus increase the whole system positioning accuracy in a large scale.Meanwhile, the error correction solution is not only applicable to same source/orbit images, but only to different source/orbit images by using theadvantage of RPC general geometric model to multi-source image processing.The paper improves the RPC model processing methods of the currentremote sensing satellite imaging, makes research on using the ground controldata to acquire correction parameters and correction system errors, updates theapplication processing method and procedure of data positioning accuracy inlarge scale, realizes instructive and objective parameter optimization, decreasessystem error interruption, increases system geometric positioning accuracy,thereby satisfy the applications in all kinds of occupations.