Research On Decipherment Model And Algorithm Of Geometric Accuracy For DEM

As the important part of the fundamental geographical information data,Digital Elevation Model(DEM)is indispensable resource in the national economic development and national defense construction,which makes its security research essential.The relevant domestic laws and regulations have confidential clear indicators for fundamental geographical information data sharing.The secret fundamental geographical information data need to be decrypted before sharing.This paper analyses polynomials,support vector machines,Gaussian processes and neural network models from precise control point transformation,smoothness and continuity,and confidentiality through analyzing the characteristic of DEM and the demand of decryption.On the basis of the existing models analyzed,this paper proposes global geometric accuracy decryption and recovery model based on radial basis functions(RBF),and a local geometric accuracy decryption and recovery model based on compactly supported radial basis functions(CSRBF)and recovery model.Finally,the decryption software is developed to verify the global decryption model with good controllability,reversible model,good data availability,high model security,and local decryption model with local compactly support,smooth continuous,precise control point transformation,which is applicable to geometric accuracy decryption.The main research contents and conclusions are as follows:(1)Research of the basis of DEM geometric precision decryption.Through analyzing the characteristic of DEM and the demand of decryption,research on basic principles of polynomials,support vector machines,Gaussian processes and neural network models,and analyze their advantages and disadvantages by contrast,which lay the foundation of decryption model research of this paper.(2)Research on DEM global geometric accuracy decryption.Establish DEM global geometric accuracy decryption and recovery model based on radial basis functions(RBF).Use the parameters of the model as the iterative factor,design the error control iteration algorithm in decryption,and finally create the variants of decryption transformation production algorithm by using chaotic model.(3)Research on DEM local geometric accuracy decryption.Establish DEM local geometric accuracy decryption and recovery model based on compactly supported radial basis functions(CSRBF)based on the characteristic of CSRBF and the demand of local decryption.And research on the effect of compactly supported radius on the decryption model and the computing method of the compactly supported radius.(4)System implementation of decryption and restoration and experimental analysis.Design and develop system to realize global and local geometric accuracy decryption and restoration for DEM.The effects of two kinds of decryption models are analyzed from precision control,model reversibility,data availability,and model security by experiments.The results show that the proposed two kinds of decryption models can better meet the needs of decryption.