Automatic 3D-visualization of a forest area based on semantic segmentation of tree crowns with height determination

№3 (2026)

Martyn K.A., Vagizov M.R., Martyn I.A., Istomin E.P., Petrov Y.A.

УДК 550.837
https://doi.org/10.47148/1609-364X-2026-3-54-60

AbstractAbout the AuthorsReferences
This article presents a study devoted to the development of a methodology for automatic 3D visualization of a forest area based on the creation of a Python script integrated into Blender. The paper describes in detail the implementation of the script in Blender, from scene space preparation with the conversion of coordinates from pixels to the global geographic coordinate system, to the placement of each individual detected tree taking into account its height. Special attention is paid to preserving height dimensions during 3D visualization.
It is shown that 3D tree templates (birch, aspen, pine, spruce, alder, deadwood) can be automatically placed according to their coordinates and height using data from a CSV-file. Testing of the program on a small area of an aerial image containing about 15 trees demonstrated a model construction time of approximately 5 minutes. The result of the developed methodology can serve as a basis for creating digital twins of forest territories.

Kristina A. Martin
Assistant at the Department of Information Systems
and Technologies
Saint-Petersburg State Forest Technical University
5, litera U, Institutskiy Lane, St. Petersburg, 194021, Russia
Senior Lecturer at the Department of Information Technologies and Security Systems
Russian State Hydrometeorological University
79, Voronezhskaya Str., St. Petersburg, 192007, Russia
е-mail: martynchris@mail.ru
ORCID: 0009-0004-3560-4821
SPIN-code: 7779-6962
AuthorID: 1187677

Marsel R. Vagizov
Doctor of Technical Sciences, Associate Professor
Vice-Rector for Information Technologies
and Digital Transformation
Russian State Hydrometeorological University
79, Voronezhskaya Str., St. Petersburg, 192007, Russia
е-mail: bars-tatarin@yandex.ru
ORCID: 0000-0003-4848-1619
Scopus ID: 57194441570
ResearcherID: I-7074-2016
SPIN-code: 4811-8943
AuthorID: 861495

Irma A. Martin
Candidate of Technical Sciences
Associate Professor of the Department of Applied Informatics
Russian State Hydrometeorological University
79, Voronezhskaya Str., St. Petersburg, 192007, Russia
е-mail: irma_martyn@mail.ru
ORCID: 0000-0002-4332-7308
Scopus ID: 57217862251
SPIN-code: 9386-0716
AuthorID: 1004119

Evgeny P. Istomin
Doctor of Technical Sciences, Professor,
Director of the Institute of Information Systems
and Geotechnologies
Russian State Hydrometeorological University
79, Voronezhskaya Str., St. Petersburg, 192007, Russia
e-mail: biom220@bk.ru
ORCID: 0000-0001-6247-4373
Scopus ID: 56951051300
SPIN-code: 6404-9070
AuthorID: 333123

Yaroslav A. Petrov
Candidate of Technical Sciences, Associate Professor,
Associate Professor of the Department of Applied Informatics
Russian State Hydrometeorological University
79, Voronezhskaya Str., Saint-Petersburg, 192007, Russia
е-mail: yaroslav.petrov025@gmail.com
ORCID: 0000-0002-9185-441X
SPIN-code: 4170-3003
AuthorID: 899415

1. Martyn Ch.A., Martyn I.A., Istomin E.P., Petrov Ya.A. Development and Verification of a YOLO‑Based Tree Detection Method. Geoinformatika. 2026;(1):112–122. DOI: 10.47148/1609-364X-2026-1-112-122.
2. Vagizov M., Potapov A. Machine Learning Technologies and Taxation Data Processing for Visualization of Geoinformation Modeling of Forest Ecosystems (Part 2). Information and Space. 2023;(1):148–156.
3. Usoltsev V.A. Prospects of 3D‑Modelling of the Spatial Structure of Forest Biomass. Eko-potencial. 2014;(2):55–71.
4. Blender documentation. Available at: https://docs.blender.org/ (accessed 03.09.2025).
5. Autodesk: 3D design, Engineering & Construction Software. Available at: https://www.autodesk.com/ (accessed 12.03.2026).
6. BlenderKit. Available at: https://www.blenderkit.com/ (accessed 16.07.2025).
7. Alekseev A.S., Nikiforov A.A. Effects of Topography on the Structure and Productivity of Forest Landscapes Using 3D Modeling in Terms of the Lisinsky Educational and Experimental Forest Enterprise. Contemporary Problems of Ecology. 2014;7(7):815-826. DOI: 10.1134/S1995425514070026.
8. Pix4D. Ground sampling distance (GSD) in photogrammetry. Available at: https://support.pix4d.com/hc/en-us/articles/202559809 (accessed 30.09.2025).

Key words: 3D-modeling; Blender; forest ecosystems; unmanned aerial vehicles (UAVs); neural networks.

Section: Geoecology