Three-dimensional parameter inversion of artificial forests through the fusion of handheld and UAV LiDAR data

Accurate and efficient extraction of forest tree parameters in plantation forests provides critical information for biomass and carbon stock estimation. This study proposes a high-precision single-tree parameter extraction method integrating airborne LiDAR and handheld LiDAR data. Using the natural forests of the Pinus koraiensis and Pinus sylvestris var. mongolica within the Hailin Forestry Bureau as experimental sites, this study employed watershed algorithms for individual tree segmentation. The investigation focused on the impact of varying resolution canopy height models (CHM) on the accuracy of single-tree delineation and parameter extraction. Concurrently, a point cloud distance clustering algorithm was utilized to achieve individual tree segmentation of normalized vegetation point clouds, with an evaluation of how different distance thresholds influence segmentation precision and parameter retrieval. Additionally, Euclidean clustering based on point cloud data was applied for single-tree segmentation, followed by the extraction of diameter at DBH parameters. The forest parameter extraction method employed in this study provides technical support for the efficient management of plantation resources.