Open-Source Framework for Creation of Canopy Height Models from UAS-Lidar Data

This work presents the integration of an open-source framework for the conformation of canopy height models. The LiDAR point cloud filtering techniques Cloth Simulation Filter and Simple Morphological Filter were evaluated, obtaining a maximum deviation or 5.77% in the generation of digital terrain models with respect to commercial tools. With the normalize digital surface model technique, a maximum difference of 7.75% was obtained from a LiDAR point cloud with a reduction factor K=10 and a GSD value of 5 cm/pix. In computational terms, a decimation process allowed to reduce the execution time, e.g., by 25% when using K=10.