Fusion of Photo with Airborne Laser Scanning

Abstract

Photogrammetry and Laser-Scanning are usually considered as complementary. Integration of these two observation methods has the potential to blend their individual advantages. The resulting benefit is likely to be higher in drone airborne mapping, which payload capacity (and thus the quality of the embedded IMU) is limited. Thus, the trajectory computed by the IMU is subject to important time-dependent errors: even if the global attitude is less adequate, it is self-coherent locally. For this reason, we propose a close integration of Photogrammetry with Laser-Scanning based on the correction of time-dependent error of the trajectory with the help of the image observations acquired by the camera. Apart from the trajectory, this hybridization requires optical correspondences between image and Laser measurements. Such full set of input data is rigorously fused together in a Bundle-Adjustment in order to better determine the trajectory, and thus the resulting point-cloud. The presented theory was practically evaluated in an airborne case against a reference solution.