基于改进YOLOv8的库尔勒香梨病虫害监测研究

打开文本图片集

Abstract： InsetpestsanddiseasesofKorlafgrantpearschasblacktardisease，backblotchdsease，otdisease，andGaolita molesta，poseseriousthreatstotheyieldandqualityofthefruit.Toenhancetheaccuracyfinsetpestsanddiseasesdetectionand effectivelycontrolthesessues，tisstudydevelopedalghtweightinsectpestsanddiseasesdetectionmodelforKorlafragantpear basedonaimprovedYOLOv8.Adataset encompasingKorla fragrantpearinsectpestsanddiseasesundervarious environmental conditions was alsoconstructed.Inthis research，the convolution module of YOLOv8 was replaced with GSConv，and the VoV-GCSP module was introduced tooptimizethe network structure，formingan eficient Slim-neck architecture.This optimizationmaintainedahighrecognitionatewhilereducingthecomputationalburdenofmodelAdditionaly，theaiteture integratedthe EMA multi-scale atention mechanismtoenhancetheabilityofmodeltorecognize features ofKorla fragrant ear insect pests anddiseases.Experimentalresults demonstrate thatthe improvedYOLOv8 modelachievedanaverage precision （mAP）of 87.9% indetecting insect pests and diseases inKorla fragrant pear，whichwasimproved compared withtheoriginal model. The modelsizewasreducedto5.5MB.Thepreciseidentificationcapabilityofthismodelcanproviderobusttechnicalsupportfor the detection ofpests and diseases in Korla fragrant pear.

Key Words： Insect pests and diseases detection; YOLOv8; Deep Learming; Korla fragrant pear; Lightweight

doi：10.13620/j.cnki.issn1007-7782.2025.01.004

中图分类号：S24;TP2 文献标识码：A

0引言

库尔勒香梨作为新疆库尔勒的特产，拥有超过2000年的栽培历史，深受国内外消费者的喜爱。（剩余4132字）