HexDEM data

This dataset comprises four airborne LiDAR point cloud samples selected to represent a broad spectrum of terrain conditions, including flat agricultural land, forested areas, mountainous terrain, and urban environments. The data originate from publicly available national and international sources, including the Dutch AHN3 program, the Polish ISOK project, and the U.S. Geological Survey 3DEP initiative.
Each dataset is provided in LAS 1.2 format and contains pre-classified point clouds with ground points identified and quality-controlled through automated classification followed by manual verification. The datasets differ substantially in spatial extent, point density, and elevation variability, with total point densities ranging from approximately 4 to over 40 points per square metre and elevation ranges spanning from a few metres in flat areas to over 1,600 metres in mountainous terrain.
The dataset is specifically designed to support research on point cloud generalisation and digital terrain model (DTM) generation. In particular, it was used to evaluate the HexDTM algorithm under varying landscape conditions, enabling systematic assessment of how terrain complexity, point density, and elevation variability influence thinning performance and resulting model accuracy.
By covering diverse geomorphological and land-use contexts, this dataset provides a robust benchmark for testing LiDAR processing workflows, classification strategies, and surface interpolation methods. It is suitable for applications in geomatics, remote sensing, terrain modelling, and computational geosciences.