Abstract: To solve the problems of over-parameterization and low geo-referencing accuracy of rational function model (RFM), a novel method of parameter optimization based on scatter matrix and elimination transformation and a new method of remnant systematic error compensation without ground control points are proposed. The proposed parameter optimization method can resolve the ill-posed problem of RFM by rejecting all excess parameters. The systematic error compensation method introduces a new correction model with Fourier coefficients. Experimental results indicate that the performance of the proposed method with less parameters is equal to that of the conventional model which all of the 78 parameters. Moreover, the ill-posed problem is effectively eliminated and thus the stabilities of estimated parameters are improved. The systematic error compensation scheme significantly eliminates the remnant systematic error of RFM and improves the geo-referencing accuracy.