Abstract
Accurate identification of crop growth stages is crucial for Accurate identification of crop growth stages is crucial for precision agriculture and automated field management. This study designed and developed an improved Swin Transformer-based detection system for wheat growth stages, with an emphasis on real-time deployment on embedded edge devices. Specifically, we incorporate a Progressive Transfer Learning strategy to ensure robust generalization on agricultural data and introduce an Ordinal Regression Loss to effectively mitigate misclassifications in transitional growth stages. The proposed approach integrates a hierarchical Transformer backbone with an optimized deployment pipeline for NVIDIA Jetson Orin NX, supporting gallery images, video streams, and live camera inputs. Experimental evaluation demonstrated that the system achieves consistently high recognition accuracy (above 93%) while maintaining real-time performance (above 12 FPS) under different modes, with moderate power consumption (6–8 W). Compared with baseline CNNs (ResNet-50, MobileNetV3) and Transformer models (ViT), the proposed design achieves a favorable balance among accuracy, efficiency, and robustness.These results suggest that the system can contribute to the development of practical agricultural monitoring and provide a step toward intelligent control strategies in precision farming.
Abstract from the paper’s OpenAlex record; open copy licensed cc-by-nc-nd. Read at the source ↗
The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.
Claims 4
Claim · model-written summary
The wheat growth stage detection system achieves recognition accuracy above 93%.
- Taxon
- wheat
Triticum aestivumt0 - Outcome
- recognition accuracye1
- Value
- above 93%e1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
t0“wheat growth stage detection”
e1“consistently high recognition accuracy (above 93%)”
Claim · model-written summary
The wheat growth stage detection system maintains real-time performance above 12 FPS.
- Taxon
- wheatTriticum aestivumt0
- Outcome
- real-time performancee1
- Value
- above 12 FPSe1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
t0“wheat growth stage detection”
e1“real-time performance (above 12 FPS)”
Claim · model-written summary
The system draws moderate power consumption of 6–8 W.
- Taxon
- wheatTriticum aestivumt0
- Outcome
- power consumptione1
- Value
- 6–8 We1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
t0“wheat growth stage detection”
e1“moderate power consumption (6–8 W)”
Claim · model-written summary
Compared with baseline CNNs and Transformer models, the proposed design achieves a favorable balance among accuracy, efficiency and robustness.
- Taxon
- wheatTriticum aestivumt0
- Outcome
- balance among accuracy, efficiency, and robustnesse1
- Direction
- favorablee1
- Comparator
- baseline CNNs (ResNet-50, MobileNetV3) and Transformer models (ViT)e1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
t0“wheat growth stage detection”
e1“Compared with baseline CNNs (ResNet-50, MobileNetV3) and Transformer models (ViT), the proposed design achieves a favorable balance among accuracy, efficiency, and robustness”