Monitoring The Spread of Cortaderia Selloana In Galicia and Northern Portugal. Deep Learning-Based Detection From Uav Imagery

The proliferation of Cortaderia selloana along the Pilgrim's route to Santiago de Compostela poses a significant threat to local ecosystems. In order to address this issue, this article investigates an efficient approach for identifying the presence of this invasive species. To this end, UAV-based imagery combined with artificial intelligence techniques are employed. As a result, an AI-based detection pipeline is designed. The proposed detection approach evaluates deep learning–based reference object detection architectures, including YOLO variants and Faster R-CNN, achieving an F1-score of up to 0.84, an [email protected] of 0.87, and an [email protected]–0.95 of 0.73, demonstrating its effectiveness in identifying this species in complex natural settings.