R scripts associated with analysis of UAS point cloud data collected at Pepperwood Preserve during 2019 field campaign with Sonoma State University/University of Oxford
Developed on a windows machine. Some scripts set up for linux cluster use. See individual script headers for additional details.
Contact Sean Reilly ([email protected]) for questions. Please open issue notes as they arise.
las_postprocessing.R: Merges Pix4D las point cloud and multispectral data, reprojects point cloud, and clips to given boundary
las_transformation.R: Performs a spatial transformation on a las point cloud based on given transformation matrix
las_plot-registration.R: Plots two las files on top of one another to visualize if registration was performed successfully
las_height-normalization.R: Height normalizes UAS and ALS point cloud data using both UAS and ALS dtm data
las_height-normalization_cluster.R: Same as above, but optimized for linux cluster (i.e. high memory capacity) use by foregoing lascatalog
csf_paramtesting_rnd1.R: Produces sequence of DTMs from a point cloud using Cloth Simulation Filter (CSF) ground finding algorithm with a supplemental NDVI reclassification filter in order to test parameter effects on DTM accuracy.
csf_paramtesting_rnd2.R: Same as above but modified to only process one parameter at a time
csf_parameter-testing_data-compile.R: Takes zonal error data from csf parameter testing and computes site-wide error values
csf_parameter-testing_visualization.R: Graphical visualization of CSF performance across parameter ranges
csf_parameter-testing_final-dtm-generation.R: Produces final DTM from a point cloud using Cloth Simulation Filter (CSF) optimized parameter set
chm_data-compile.R: Generates large dataset containing chm values, dtm values, errors, vegetation classes, topography classes, and burn severities. Also generates several reference plots.
chm_generation.R: Generates canopy height models for UAS and ALS height normalized data