Automated Stock Volume Estimation Using UAV-RGB Imagery.

Area of Science:

• Forestry Science
• Remote Sensing
• Ecology

Background:

• Forests are vital carbon sinks, with tree crown size influencing growth, carbon sequestration, and habitat.
• Accurate forest stock volume estimation is crucial for assessing land-use impacts.
• Traditional methods like Diameter at Breast Height (DBH) measurements are labor-intensive and impractical for large areas.

Purpose of the Study:

• To investigate the correlation between tree crown area and forest stock volume.
• To develop an automated empirical model for stock volume and above-ground biomass estimation using Unmanned Aerial Vehicle (UAV) RGB data.
• To enhance the efficiency and interpretability of forest assessments.

Main Methods:

• Utilized high-resolution UAV-RGB imagery for forest data acquisition.
• Developed an empirical model correlating tree crown area with stock volume.
• Validated the model using extensive training and testing sites.

Main Results:

• A strong exponential correlation was found between crown area and stem stock volume (R²=0.67, MSE=0.0015).
• The developed UAV-based model demonstrated high accuracy in estimating cumulative stock volume (R²=0.75).
• The model significantly improves the efficiency and convenience of forest stock volume assessment.

Conclusions:

• Drone technology, combined with traditional forestry techniques, offers a powerful tool for precise and efficient stock volume estimation.
• The developed empirical model effectively automates stock volume and above-ground biomass calculations.
• This approach facilitates more accurate monitoring of forest ecosystems and their role in the carbon cycle.