High performance R functions for forest inventory based on Terrestrial Laser Scanning (but not only) point clouds.


This package is a refactor of the methods described in this paper.

The algorithms were rewritten in C++ and wrapped in R functions through Rcpp. The algorithms were reviewed and enhanced, new functionalities introduced and the rebuilt functions now work upon lidR’s LAS objects infrastructure.

This is an ongoing project and new features will be introduced often. For any questions or comments please contact me through github. Suggestions, ideas and references of new algorithms are always welcome - as long as they fit into TreeLS’ scope.

Main functionalities

Coming soon:



Install TreeLS latest version

On the R console, run:


Legacy code

For anyone still interested in the old implementations of this library (fully developed in R, slow but suitable for research), you can still use it. In order to do it, uninstall any recent instances of TreeLS and reinstall the legacy version:

devtools::install_github('tiagodc/TreeLS', ref='old')

Not all features from the old package were reimplemented using Rcpp, but I’ll get there.


Example of full processing pipe until stem segmentation for a forest plot:


# open artificial sample file
file = system.file("extdata", "pine_plot.laz", package="TreeLS")
tls = readTLS(file)

# normalize the point cloud
tls = tlsNormalize(tls, keepGround = T)
plot(tls, color='Classification')

# extract the tree map from a thinned point cloud
thin = tlsSample(tls, voxelize(0.05))
map = treeMap(thin, map.hough(min_density = 0.03))

# visualize tree map in 2D and 3D
xymap = treePositions(map, plot = TRUE)
plot(map, color='Radii')

# classify stem points
tls = stemPoints(tls, map)

# extract measures
seg = stemSegmentation(tls, = 15))

# view the results
tlsPlot(tls, seg)
tlsPlot(tls, seg, map)