On the Use of Inertial/GPS Velocity Control in Sensor Calibration and Orientation

Since the 1980’s and 1990’s, GPS and INS/GPS data, respectively, have been used to support sensor calibration and orientation of kinematic primary data acquisition systems. Aerial control position and attitude observations, well-known to the photogrammetric community, are used for Direct Sensor Orientation (DSO) and combined with the traditional aerial triangulation image observations to reduce ground control and mitigate the old constraints on the shape of blocks in Integrated Sensor Orientation (ISO). In the past two years, at the Institute of Geomatics (IG), research has been conducted to improve the reliability and robustness of the mathematical models for aerial position and attitude control. This effort has already led to new models for relative position and attitude control. In this paper, the authors introduce models for the temporal calibration of multi-sensor systems. For this purpose, the INS/GPS-derived velocities are used as velocity control information. To the best knowledge of the authors, this information has not yet been used for ISO. The paper is based on two main ideas. First, that INS/GPS data include linear and angular velocities, not only position and attitude. Second, that the sensor calibration and orientation problem is not just a 3D spatial problem, but a 4D spatio-temporal problem and that, as experience has shown, the temporal dimension must be taken into account. This paper describes the proposed models in which the velocities are included to calibrate the time error of the various sensors of a multi-sensor system with respect to the accurate and precise GPS time reference frame. The authors believe that the addition of the velocities to calibrate time could make the difference between “good” and “excellent” results at no additional cost as the newly used data –the velocities– are already available from the INS/GPS processing.