Succulent plant image categorization suffers from subjectivity and limited number of available images, and these difficulties are intensified by the apparent lack of field surveys. In this study, we present a new approach to classify succulents using consumer-grade Unmanned Aerial Vehicles (UAVs) and deep learning. We applied various super-resolution algorithms with different image enhancement design strategies. In addition, we develop a new object detection model based on MambaIR and YOLOv10, achieving a mean average precision (mAP) of 0.851, outperforming other state-of-the-art object detection algorithms. The application of UAV s and super-resolution technology greatly facilitates the detection of succulents and provides a practical solution for precision crop monitoring.
UAV-Assisted Succulent Farmland Monitoring with Improved YOLOv10 and MambaIR
21.03.2025
2634823 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch