3D Tree Dimensionality Assessment Using Photogrammetry and Small Unmanned Aerial Vehicles

Area of Science:

• Forestry Science
• Computer Vision
• Remote Sensing

Background:

• Accurate 3D tree models are crucial for understanding forest dynamics, but traditional methods are limited by cost and labor.
• Existing technologies like airborne and terrestrial LiDAR offer incomplete or costly tree reconstructions.

Purpose of the Study:

• To develop an affordable and precise method for generating detailed 3D tree models.
• To overcome technological and logistical constraints in forest biometrics.

Main Methods:

• Utilized small unmanned aerial vehicles (UAVs) equipped with inexpensive cameras for image acquisition.
• Employed computer vision algorithms and open-source software for processing images into detailed 3D tree reconstructions.
• Evaluated various UAV platforms, image acquisition strategies, and processing algorithms.

Main Results:

• Successfully created a cost-effective workflow for generating precise, comprehensive 3D models of individual trees and small groups.
• Demonstrated the potential of UAV-based photogrammetry for detailed forest structure analysis.

Conclusions:

• The developed UAV-based method offers a significant advancement over traditional techniques for 3D tree modeling.
• This approach is expected to improve forest self-organization studies and individual-tree-based forest modeling.