Forest Cover Mapping with Deep Learning and OBIA
Research Paper#Remote Sensing, Deep Learning, Forest Cover Mapping🔬 Research|Analyzed: Jan 3, 2026 19:07•
Published: Dec 29, 2025 04:23
•1 min read
•ArXivAnalysis
This paper presents a novel approach, ForCM, for forest cover mapping by integrating deep learning models with Object-Based Image Analysis (OBIA) using Sentinel-2 imagery. The study's significance lies in its comparative evaluation of different deep learning models (UNet, UNet++, ResUNet, AttentionUNet, and ResNet50-Segnet) combined with OBIA, and its comparison with traditional OBIA methods. The research addresses a critical need for accurate and efficient forest monitoring, particularly in sensitive ecosystems like the Amazon Rainforest. The use of free and open-source tools like QGIS further enhances the practical applicability of the findings for global environmental monitoring and conservation.
Key Takeaways
- •ForCM integrates deep learning with OBIA for improved forest cover mapping.
- •The study evaluates and compares several deep learning models (UNet, UNet++, ResUNet, AttentionUNet, ResNet50-Segnet).
- •The method achieves higher accuracy than traditional OBIA.
- •The research highlights the potential of free and user-friendly tools like QGIS for environmental monitoring.
Reference / Citation
View Original"The proposed ForCM method improves forest cover mapping, achieving overall accuracies of 94.54 percent with ResUNet-OBIA and 95.64 percent with AttentionUNet-OBIA, compared to 92.91 percent using traditional OBIA."