Tree Crown Segmentation and Stock Volume Prediction Using Unoccupied Aerial Vehicle Imagery

Abstract

Forests are critical components of the global carbon cycle in which [1] tree crown size is a versatile ecological indicator that influences various factors such as tree growth, wind resistance, shading, and carbon sequestration [2]. Important aspects of tree crowns include their length, base height, and crown radius [3]. Understanding the relationship between tree crown area and stock volume is crucial for comprehending carbon storage and sequestration within forest ecosystems, and it serves as a key metric for assessing the impact of land-use changes on ecological processes [4].Traditionally, estimating stock volume through ground-based observations is labor-intensive and often impractical under certain environmental conditions [5]. Henseforth, the increasing availability of large-scale, high-resolution UAV imagery has opened new avenues for research in remote sensing and computer-based photointerpretation of trees [6] It helps improve and enhance forest studies by making the procedure time-friendly and more interpretable. While previous studies have established correlations between DBH (Diameter at Breast Height) and stock volume [7], this study aims to explore the correlation between tree crown area and stock volume using UAV data. The findings underscore a significant association, demonstrating the potential of integrating drone technology with traditional forestry techniques for more accurate and efficient stock volume estimation [8]. © 2024 IEEE.