sylva-rs 2.0.0

Terrestrial and airborne laser scanning processing for forest ecology and remote sensing research.
Documentation

sylva-rs

crates.io PyPI

Terrestrial and airborne laser scanning processing for forest ecology: the Rust core of Sylva, which also ships as the sylva-rs Python package (pip install sylva-rs, then import sylva) and the sylva command (cargo install sylva-cli).

cargo add sylva-rs

The mta-model feature (on by default) compiles in the trained model that resolves RIEGL multiple-time-around ambiguities when reading .rxp.

  • PointCloud with typed attributes; LAS/LAZ, PLY, text, mesh, GeoTIFF and RIEGL .rxp / .rdbx I/O; coordinate transforms and reprojection
  • filters, k-d tree queries, clustering, PCA normals, interpolation, masks
  • ground classification (CSF, PMF), DTM, CHM
  • stem detection, DBH, tree segmentation, crown metrics
  • registration (Kabsch, ICP) and marker-free coregistration of scan positions
  • cylinder QSMs, leaf / wood separation and leaves on QSMs
  • Shots: pulse data with misses, stored as Parquet and streamed
  • ray-traced voxel grids with FPL / PPL / transmittance attenuation and plant, leaf and wood area density
  • airborne lidar over tiled areas: metrics, trees, canopy from the trajectory, full waveforms; TLS and ALS fusion; change detection between surveys
use sylva_rs::{ground, io, trees};

let cloud = io::read("plot.laz")?;
let mask = ground::csf_ground_mask(&cloud.xyz, &ground::CsfParams::default());
let ground_pts: Vec<_> = cloud.xyz.iter().zip(&mask).filter(|(_, &g)| g).map(|(p, _)| *p).collect();
let dtm = ground::make_dtm(&ground_pts, 0.5, None)?;
let heights = ground::heights_above(&cloud.xyz, &dtm);
let stems = trees::detect_stems(&cloud.xyz, &heights, &trees::StemParams::default());

Documentation: https://uq-eorc.github.io/Sylva/. Licence: GPL-3.0-or-later. To cite Sylva, see the citation in the repository.