Integrating UAV-SfM and Airborne Lidar Point Cloud Data to Plantation Forest Feature Extraction

A low-cost but accurate remote-sensing-based forest-monitoring tool is necessary for regularly inventorying tree-level parameters and stand-level attributes to achieve sustainable management of timber production forests. Lidar technology precise multi-temporal data collection expensive. UAV-based optical sensing method an economical flexible alternative collecting high-resolution images generating point cloud orthophotos mapping lacks height accuracy. This study proposes a protocol integrating UAV equipped without RTK instrument airborne lidar sensors (ALS) characterizing tree stand use in plantation forest management. The proposed primarily relies on the ALS-based digital elevation model (ALS-DEM), structure-from-motion technique generated surface (UAV-SfM-DSM), their derivative canopy (UAV-SfM-CHM). Following traditional inventory approaches, few middle-aged mature stands Hinoki cypress (Chamaecyparis obtusa) forests were used investigate performance via model. Results show that can improve UAV-SfM referencing transformation With derived CHM data, this estimate with RMSE ranging from 0.43 m 1.65 m, equivalent PRMSE 2.40–7.84%. estimates between approaches are highly correlated (R2 = 0.98, p < 0.0001), similarly, annual growth rate (HAGR) also significantly 0.78, 0.0001). percentage HAGR trees behaves as exponential decay function over 8-year period. density, volume stocks, basal area, relative spacing error less than 20% both approaches. Intensive regular thinning helps retain clear crown shape feature, therefore, benefitting segmentation deriving attributes.