农业装备与车辆工程2025，Vol.63Issue(11)：15-23,9.DOI:10.3969/j.issn.1673-3142.2025.11.003
基于改进YOLOv11n的轻量化植物病虫害检测算法
Lightweight plant disease detection algorithm based on improved YOLOv11n
摘要
Abstract
In recent years,plant diseases and insect pests have occurred frequently and shown an intensifying trend.The use of target detection algorithms to achieve intelligent monitoring and control has become an efficient approach.However,traditional methods usually require transmitting collected images back to ground-based computing centers for processing,which was limited by high computing costs and great implementation difficulties.To address this issue,a lightweight target detection algorithm suitable for edge devices was proposed.Firstly,this method adopted the efficient feature extraction structure Starnet as the backbone network,which significantly reduced the number of parameters while extracting more discriminative image features.Secondly,the C3k2_SECA module was introduced in the Neck part to enhance the multi-scale feature fusion capability.Finally,the original Head structure was replaced with the LightHead-SBN detection head:the computational efficiency was improved through a convolution sharing mechanism,the feature independence was maintained by means of independent BN layers,and the DFL decoding module was retained to ensure high-quality target regression performance.Experiments conducted on the self-built disease and pest dataset RPH showed that,compared with YOLOv11n,the improved model achieved a 31.3%reduction in the number of parameters,a 22.2%decrease in floating-point operations,a 5.9%increase in precision(equivalent to a 7.2%relative improvement),and a 31.7%enhancement in reasoning speed.In conclusion,the proposed algorithm significantly reduced the demand for computing resources while improving detection accuracy and inference efficiency.It provides a feasible solution for the practical deployment of edge devices such as unmanned aerial vehicles(UAVs)and has broad application prospects in the field of intelligent prevention and control of agricultural diseases and insect pests.关键词
目标检测/改进YOLOv11n/病虫害/轻量化Key words
object detection/improved YOLOv11n/pest and disease/lightweight分类
信息技术与安全科学引用本文复制引用
石子豪,刘延旭..基于改进YOLOv11n的轻量化植物病虫害检测算法[J].农业装备与车辆工程,2025,63(11):15-23,9.基金项目
德州学院校级科研项目资助(2022XJRC111) （2022XJRC111）
