Optimizing georeferencing of airborne survey systems by INS/DGPS

This research scrutinizes the method of INS/DGPS integration to provide the parameters of exterior orientation needed for direct georeferencing of airborne imagery more reliably and with better accuracy. Since the well established method of INS-GPS data integration by means of Kalman filtering suppresses sensor noise over a limited frequency bandwidth only, a new filtering methodology has been developed to optimally eliminate noise over the whole motion band of interest. The proposed method includes optimal band-limiting for the inertial signal and de-noising by use of wavelets. The latter accommodates a model for short-term inertial errors, especially applicable to ring-laser gyro and fibre optic gyro technologies and presents a technique for adaptive estimation of the model parameters based solely on data analysis. In the second part of the thesis, the influence of inertial measurements on the process of GPS ambiguity determination is investigated with the purpose of optimizing INS/DGPS integration in this respect. An analytical proof demonstrating in which cases data integration by means of a centralized Kalman filter provides considerable advantages over its de-centralized counterpart is presented, and the contribution of inertial systems of different accuracy classes is evaluated. In addition, the operational aspects of airborne mapping with INS/DGPS are analyzed and strategies for minimizing the effect of hardware integration errors on the process of direct georeferencing are proposed. Lastly, the feasibility of using INS/DGPS data for the derivation of the parameters of exterior orientation is demonstrated with respect to aerotriangulation without ground control.