基于改進(jìn)YOLOv8s模型的河蟹幼苗雌雄檢測方法

打開文本圖片集

GenderDetectionMethodforCrabsSeedlingsBasedonEnhancedYOLOv8sModel

LI Xian，MA Ming，HUZhi-gangetal（ScholofMchanicalEnginering，WuhanLightIndustry University，Wuhan，Hubei0496） AbstractInespsetoteproblmofuceardierentiatiobetwenalesndealesuringthsdingsageofiverabfi cientaccuracyinmanalclasificatioofaleandfemaleinvrabsdling.Tisticleproposametdfordetetigtederf rivercrabsedingasedonanimprovedOLO8odel.eiprovementmetodisasflow：fistlyeplacethfourthlayeCfodule inthebackboneetworkwithCfGAM（globalatentionchanism，GAM）module，djust hewightoffatureiforatioandducete lossoffeaturefoatiodingtassidlyplacealdarysfciiheU（edtef sequentialevideceforintersetooveruiov）boundysfuctiotohaethancroxualityduringtepreditoprosd improvetheodel’sneralzatiobilityfiallearestigoriteplatiousaplingtodineadetwokisplacdih CARAFE（contentawarereassemblyoffeatures）upsaplingmetodichivesteodellargereceptivefeldndimproesitsfo ance.The experimental results sow that the accuracy，recall，and average precision of the improved model are98. 4% ，91. 1% ，and 96. 1% ， respctivelyhcre3..d2.9prenaeoterthaeigialodel.eulsicatethaslityoflnga chine vision to the clasification of male and female crab seedlings and the effectiveness of the improved method.

Key wordsCrab seedlings；YOLOv8s；WIoU；GAM；CRAFF

近年來，隨著人工智能技術(shù)的不斷進(jìn)步，越來越多的行業(yè)為了追求高質(zhì)量發(fā)展，將傳統(tǒng)技術(shù)與人工智能技術(shù)進(jìn)行結(jié)合[1-2]。（剩余8718字）