YOLOv10 and Mamba-Based Super-Resolution for Smart Rose Growth Monitoring Using UAV Imagery

To meet market demand and optimize production efficiency, accurately assessing the growth status of fresh-cut flowers is crucial. Due to time and labor constraints, manually assessing cultivation conditions by growth stage is challenging. This study introduces a novel method that integrates UAV technology with deep learning to enhance rose cultivation in greenhouses. By leveraging Super-Resolution Reconstruction (SRR) algorithms and advanced target detection techniques, the method significantly improves image quality and accuracy in assessing rose growth. The approach contributes to agricultural technology advancements and is expected to optimize productivity and quality across various crops. Notably, the methodology achieves a notable mAP (Mean Average Precision) score of 0.872.