J. For. Sci., 2025, 71(10):501-515 | DOI: 10.17221/53/2025-JFS

The use of LiDAR for the documentation and modelling of cultural heritage sites hidden by the forest canopyOriginal Paper

Nikola Žižlavská ORCID..., Stanislav Herber ORCID...

Department of Forest Management and Applied Geoinformatics, Faculty of Forestry and Wood Technology, Mendel University in Brno, Brno, Czech Republic

The large number of charcoal kiln sites (CHKS) preserved as cultural heritage monuments demonstrates how extensive forest management for charcoal production has influenced the present forest dynamics and contributed to industrial expansion. Therefore, it is necessary not only to have a reliable methodology for detecting and documenting individual kiln sites for management and protection purposes but also to present the results in a meaningful way to the public. The aim is to optimise the data processing workflow from airborne laser scanning (ALS) point cloud to printable model (from LAS format to STL), determine the influence of vegetation cover at the time of data collection on the quality of the resulting model, verify the quality of printed models using photogrammetry, and finally, produce printed models of CHKS as cultural heritage objects in a form that can be effectively presented to the public. After comparison of different ground filtering methods, we conclude that the most accurate method for creating a precise ground representation for our area of interest was the Cloth Simulation Filtering (CSF) algorithm. From the filtered point cloud, a high-resolution raster surface was generated, which served as the basis for CHKS detection. Using our proposed methodology – combining the topographic position index (TPI) with a 0–5% slope threshold – we achieved a significant improvement in detection performance compared to using a zero-slope threshold alone, with the F1 score increasing from 0.667 to 1.000. Subsequently, the most representative kiln site was selected, which was then processed and optimised using various software tools, resulting in an exchangeable STL file suitable for dissemination and 3D printing. The accuracy of the printable model was then evaluated. We conclude that point cloud from the winter flight campaign achieved higher accuracy. The average height differences were similar; however, the spatial distribution varied between the two clouds. The model from the winter flight campaign had more evenly distributed deviations and overall was better for modelling the CHKS.

Keywords: 3D mesh; 3D objects; 3D printing; charcoal kiln sites; photogrammetry

Received: July 15, 2025; Revised: September 5, 2025; Accepted: September 17, 2025; Published: October 30, 2025  Show citation