During the last five years different research institutes and private companies stared to implement new algorithms to analyze and extract features from LiDAR data but only a few of them also created a public available software. In the field of forestry there are different examples of software that can be used to extract the vegetation parameters from LiDAR data, unfortunately most of them are closed source (even if free). In 2014 we started the development of the library LESTO (LiDAR Empowered Sciences Toolbox Opensource): a set of modules for the analysis of LiDAR point cloud with an Open Source approach with the aim of improving the performance of the extraction of the volume of biomass and other vegetation parameters on large areas for mixed forest structures.
LESTO contains a set of modules for data handling and analysis implemented within the JGrassTools spatial processing library. The main subsections are dedicated to 1) preprocessing of LiDAR raw data mainly in LAS format (utilities and filtering); 2) creation of raster derived products; 3) flight-lines identification and normalization of the intensity values; 4) tools for extraction of vegetation and buildings.
The core of the LESTO library is the extraction of the vegetation parameters following the single tree based approach. The implemented algorithms work on both LiDAR derived raster datasets (DTM, DSM) or raw point clouds. The methods range between the simple extraction of tops and crowns from local maxima, the region growing method, the watershed method and individual tree segmentation on point clouds.
The information resulting from the analysis of the morphology and vegetation for a river basin can be used to evaluate the possible hazard along the stream network. In particular we investigated the problem of Large Wood recruitment and transportation along the channel network in mountain basins using a simplified methodology.
LESTO and the modules for Large Wood debris evaluation are integrated in the JGrassTools project and are available for download at www.jgrasstools.org. A simple and easy to use graphical interface to run the models is available at https://github.com/moovida/STAGE/releases.
