GPS-based attitude determination for airborne remote sensing

Abstract

The design and testing of a GPS based attitude determination system for the direct georeferencing of airborne imagery is discussed. By combining dual-frequency, high data rate GPS receivers with a relatively simple gyroscopic attitude and heading reference system, position and orientation data can be collected at a rate which matches that of a typical airborne scanner, with an accuracy adequate for georeferencing imaged features to a few metres.

Data collected in two flight trials in October 1998 have been processed using software developed at University College London and the University of Newcastle upon Tyne, to determine a best estimate of the aircraft trajectory. The software employs single epoch ambiguity resolution, a modified double-differencing algorithm for direct determination of attitude parameters, and a Kalman filter for data integration. The method presented here uses a robust and efficient routine for ambiguity resolution, and exploits the additional redundancy of the modified double-differencing algorithm to detect and remove any poor observations from the final solution. Details of the computation methods are presented, together with an analysis of the results.