metrics_from_las(): recomputes tree- and
plot-level metrics (DBH, height, crown base height, plot statistics)
from an already classified point cloud (LAS with ASPRS
classes ground = 2, understory = 3, wood = 4, crown = 5), without
repeating the full Forest_seg() segmentation.pic_analyze_cloud(): pre-processing
diagnostic tool that assesses point-cloud quality and structure (point
density per m^2, voxel density distribution), returns plots for
interactive use and optionally exports a PDF report. The loaded cloud
stays in memory for a subsequent Forest_seg() run.Forest_seg() plot report now includes
crown_volume_m3 and understory_volume_m3,
computed from the final classified points (count of occupied voxels at
canopy_vox_dim resolution times the voxel volume) and
rounded to whole cubic metres.cbh_max_gap_layers is retained for backward compatibility
but is no longer used.cls == 0), not only
rejected clusters;h_trunk_factor is clamped to [0.3, 1.0],
so rescue is never stricter than the base pass on height;minPts is halved, so sparse noise can
re-cluster;output_format: LAS is the default; any value other than
"xyz" falls back to LAS..) as decimal separator for
international use; crown/understory volumes are written as whole cubic
metres..analyze_canopy_float(), which was never wired into the
pipeline..assign_components_bfs_float().H_threshold value;
it now falls back to dav_understory_max_start..lor_create_voxels() is marked internal
and no longer appears in the package help index.Updated defaults for Forest_seg():
N = 3000h_rescue_max = 5dbh_max_radius = 0.5canopy_min_density = 500dav_understory_max_start = 1.3dav_eps = 2dav_minPts = 4dbh_heights = c(1.3, 1.8, 2.3) (fixed internal
parameter)