Abstract
Pest infestation is one of the primary causes of decreased cotton yield and quality. Rapid and accurate identification of cotton pest categories is essential for producers to implement effective and expeditious control measures. Existing multi-scale cotton pest detection technology still suffers from poor accuracy and rapidity of detection. This study proposed the pruned GBW-YOLOv5 (Ghost-BiFPN-WIoU You Only Look Once version 5), a novel model for the rapid detection of cotton pests. The detection performance of the pruned GBW-YOLOv5 model for cotton pests was evaluated based on the self-built cotton pest dataset. In comparison with the original YOLOv5 model, the pruned GBW-YOLOv5 model demonstrated significant reductions in complexity, size, and parameters by 68.4%, 66.7%, and 68.2%, respectively. Remarkably, the mean average precision (mAP) decreased by a mere 3.8%. The pruned GBW-YOLOv5 model outperformed other classic object detection models, achieving an outstanding detection speed of 114.9 FPS. The methodology proposed by our research enabled rapid and accurate identification of cotton pests, laying a solid foundation for the implementation of precise pest control measures. The pruned GBW-YOLOv5 model provided theoretical research and technical support for detecting cotton pests under field conditions. © 2024 Society of Chemical Industry.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
