ODYN: AN ONLINE DYNAMIC NETWORK SOLVER FOR PHOTOGRAMMETRY AND LIDAR GEO-REFERENCING
Keywords: Dynamic network, Sensor orientation, LiDAR, Photogrammetry, IMU
Abstract. In recent years, many methods have been proposed to improve the quality of the estimated trajectories for airborne or terrestrial mapping platforms leveraging multi-sensor fusion. One of the motivating applications is, for example, the use laser scanners on small unmanned aerial vehicles, where the typically employed low-cost MEMS inertial sensors do not allow for satisfactory direct geo-referencing of the laser points. In this work we introduce ODyN, an online Dynamic Network solver that can fuses information from cameras, GNSS and inertial sensors in a single adjustment. It can be employed to estimate a high-frequency trajectory for precise direct geo-referencing, to improve photogrammetric reconstructions in challenging scenarios or to determine several types of system calibration parameters. The presented solver is hosted by the University of Geneva and is free for anybody to use. In this work we present a use case in airborne mapping where the obtained trajectory estimates are improved with respect to the recursive smoothing approaches conventionally used in direct geo-referencing.