The phenotypic characteristics of pomegranates at different growth stages significantly influence yield, making real-time detection essential for optimizing production. Additionally, complex environmental conditions complicate this process. Current research on pomegranate growth stage detection is scarce, with many studies ignoring environmental interference. To address these gaps, this study proposes a hybrid and efficient scale-aware pomegranate detection transformer model (HESP-DETR) based on Real-Time DEtection TRansformer (RT-DETR). Firstl...
