基于改进YOLOv5s算法的草莓病虫害检测研究  

Research on Strawberry Disease and Pest Detection Based on Enhanced YOLOv5s Algorithm

作　　者：马赛 丁健 张火强 汪慧 Ma Sai;Ding Jian;Zhang Huoqiang;Wang Hui(School of Advanced Manufacturing Engineering,Hefei University,Hefei,Anhui 230601,China)

机构地区：[1]合肥大学先进制造工程学院,安徽合肥230601

出　　处：《黑龙江工业学院学报（综合版）》2024年第3期94-100,共7页Journal of Heilongjiang University of Technology(Comprehensive Edition)

基　　金：2023年安徽省大学生创新创业训练计划项目(项目编号:S202311059261X)。

摘　　要：草莓的生长容易受到多种病虫害的影响,为了快速、准确地检测草莓植株在生长过程中所受病虫害的情况,通过采集草莓生长过程中的图像信息,应用YOLOv5s算法进行分析处理,将CBAM注意力机制集成到YOLOv5s模型中,用以增强模型的感受力和表达能力,通过选择SIOU损失函数替代GIOU损失函数进一步加快模型收敛。研究结果表明,经过改进的算法模型准确率达到了95.2%,召回率提升至97.2%,平均精度均值提高至98.5%。一定程度上满足草莓病虫害的检测。Strawberry growth is susceptible to a variety of diseases and pests.In order to quickly and accurately detect the diseases and pests suffered by strawberry plants during the growth process,this paper collects image information during the growth process of strawberry,applies YOLOv5s algorithm for analysis and processing,and integrates CBAM attention mechanism into YOLOv5s model.In order to enhance the sensibility and expressiveness of the model,SIOU loss function is selected to replace GIOU loss function to accelerate the convergence of the model.The results show that the accuracy of the improved algorithm model reaches 95.2%,the recall rate increases to 97.2%,and the average accuracy increases to 98.5%.To some extent,it can satisfy the detection of strawberry pests and diseases.

关 键 词：病虫害检测 YOLOv5s CBAM SIOU损失函数 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]