YOLO-Wheat: A More Accurate Real-Time Detection Algorithm for Wheat Pests

Yongkang Liu, Qinghao Wang, Qi Zheng and Yong Liu ()
Additional contact information
Yongkang Liu: College of Electronic Engineering, Heilongjiang University, Harbin 150080, China
Qinghao Wang: College of Electronic Engineering, Heilongjiang University, Harbin 150080, China
Qi Zheng: College of Electronic Engineering, Heilongjiang University, Harbin 150080, China
Yong Liu: College of Electronic Engineering, Heilongjiang University, Harbin 150080, China

Agriculture, 2024, vol. 14, issue 12, 1-17

Abstract: As a crucial grain crop, wheat is vulnerable to pest attacks throughout its growth cycle, leading to reductions in both yield and quality. Therefore, promptly detecting and identifying wheat pests is essential for effective pest management and to guarantee better wheat production and quality. Wheat pests exhibit considerable diversity and are often found in complex environmental contexts. Intraspecies variation among wheat pests can be substantial, while differences between species may be minimal, making accurate pest detection a difficult task. We provide an enhanced algorithm, YOLO-Wheat, based on YOLOv8, to solve the aforementioned issues. The proposed YOLO-Wheat, an extension of YOLOv8, integrates SimAM into the C2f module to enhance feature extraction capabilities. Additionally, a novel feature fusion technique, CGconcat, is introduced, which enhances fusion efficiency by applying channel weighting to emphasize critical feature information. Moreover, the EMA attention mechanism is implemented before the detection head to preserve feature information through multipath processing, thereby addressing detection challenges posed by pests of varying sizes. Experiments revealed that YOLO-Wheat achieved an mAP@0.5 of 89.6%, reflecting a 2.8% increase compared to its prior performance. Additionally, mAP@0.5:0.95 reached 46.5%, marking a 1.7% improvement. YOLO-Wheat also performs better than other popular object detection algorithms (YOLOv5, YOLOv10, RT-DETR), and the model is successfully deployed for simple real-time detection. These results demonstrate that YOLO-Wheat can achieve real-time high-precision detection for wheat pests.

Keywords: wheat pest detection; YOLOv8; attention mechanism; real-time object detection (search for similar items in EconPapers)
JEL-codes: Q1 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17 Q18 (search for similar items in EconPapers)
Date: 2024
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2077-0472/14/12/2244/pdf (application/pdf)
https://www.mdpi.com/2077-0472/14/12/2244/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jagris:v:14:y:2024:i:12:p:2244-:d:1538996

Access Statistics for this article

Agriculture is currently edited by Ms. Leda Xuan

More articles in Agriculture from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().