The Integration of Uavand Backpack Lidar Systems for Forest Inventory

摘要

Light detection and ranging (Lidar) has been demonstrated with strong capability in capturing three-dimensional (3D) forest structures. The reducing cost and weight of lidar sensors and the development of positioning systems make light-weighted near-surface lidar platforms, such as unmanned aerial vehicle (UAV) and backpack, become possible. These light-weighted lidar platforms have great potentials to be used as complimentary tools to improve both the accuracy and efficiency of forest inventory. In this study, we presented our self-developed low-cost UAV lidar system and backpack lidar system. A kinetic calibration algorithm and an improved 3D simultaneous localization and mapping algorithm were used to improve the positioning accuracy of the UAV lidar system and backpack lidar system, respectively. We then explored the possibility to automatically register backpack lidar data with UAV lidar data in a conifer forest using a two-step method based on segmented individual tree locations and the iterative closest point algorithm. The results of this study can provide guidance for a systematic lidar solution in forest inventory.