DEEP LEARNING BASED AERIAL IMAGERY CLASSIFICATION FOR TREE SPECIES IDENTIFICATION

Abstract

Forest monitoring and tree species categorization has a vital importance in terms of biodiversity conservation, ecosystem health assessment, climate change mitigation, and sustainable resource management. Due to large-scale coverage of forest areas, remote sensing technology plays a crucial role in the monitoring of forest areas by timely and regular data acquisition, multi-spectral and multi-temporal analysis, non-invasive data collection, accessibility and cost-effectiveness. High-resolution satellite and airborne remote sensing technologies have supplied image data with rich spatial, color, and texture information. Nowadays, deep learning models are commonly utilized in image classification, object recognition, and semantic segmentation applications in remote sensing and forest monitoring as well. We, in this study, selected a popular CNN and object detection algorithm YOLOv8 variants for tree species classification from aerial images of TreeSatAI benchmark. Our results showed that YOLOv8-l outperformed benchmark’s initial release results, and other YOLOv8 variants with 71,55% and 72,70% for weighted and micro averaging scores, respectively.