Summary
This paper applies YOLOv10, a state-of-the-art deep learning object detection framework, to automated barley head detection from high-resolution UAV imagery. Using a dataset of UAV-captured images supplemented with the Global Wheat Head Dataset, the approach achieves a mean Average Precision of 0.83 at IoU 0.5, establishing a new benchmark for barley head detection. The work advances precision agriculture by enabling rapid, non-destructive crop monitoring and yield estimation.
Regional applicability
The methodology is geographically agnostic and could be adapted to barley production systems in the United Kingdom and elsewhere. Transferability would depend on validating the trained model against UK barley varieties and growing conditions, particularly given differences in head morphology and UAV altitude protocols.
Key measures
Mean Average Precision (mAP) at Intersection over Union (IoU) 0.5; barley head detection accuracy from UAV-captured imagery.
Outcomes reported
The study demonstrates automated detection of barley heads from high-resolution UAV imagery using YOLOv10, achieving a mean Average Precision of 0.83 at IoU 0.5. The work establishes a benchmark for barley head detection to support yield estimation and crop monitoring applications.
Topic tags
Dig deeper with Pulse AI.
Ask about this record, its theme or its relevance to UK farming and policy. Pulse AI uses selected catalogue evidence and cites the sources it draws on.