一种基于图像识别的小麦分选系统

打开文本图片集

中图分类号：TN911.73-34；TP23 文献标识码：A 文章编号：1004-373X（2025）15-0158-07

Wheat sorting system based on image recognition

HUYan1²，ZHAOYuhong'，ZHANG Hongming²，LUBo3，YINXianghui’，LIXiaoxing1，²，HUKui²，FANYu，FUJia². WANGFudi²，ZENGChao，SUNBo¹，SHENYongcai4，DAIShuyu5 （1.SchoolofElectrical Engineering，Universityof South China，Hengyang421o1，China; 2.InstituteofPlasma Physics，HefeiInstitutesofPhysicalScience，ChineseAcademyofSciences，Hefei 230o31，China; 3.Science IslandBranch，GraduateSchoolofUniversityofScienceandTechnologyofChina，Hefei230o31，China; 4.School of Physicsand Materials Engineering，Hefei Normal University，Hefei 23O6O1，China; 5.SchoolofPhysics，DalianUniversityofTechnology，Dalian116O24，China）

Abstract：Under the background thatdep learning technologyisincreasinglypromoting theimprovementofthedetection accuracyofwheatimperfectgrains，acompletesortingcontrolsystemisdevelopedinordertorealizethefectivecontroland dockingfromidentificationtoscreening.ThesystemmainlyusesthestrategyofcombiningimagerecognitionandFPGAreal-time control，prosseatediossofgatidgdptei control.Thesystemisbasedonafullyautomaticcontrolprocess，anditsuccessfullyrealizestheeficientsortingoerationof wheatunderthepremiseofensuringaccuratecontroloftime.Afterthetestof1OOwheat grainsforeachofthefivetraits，itwas foundthattheidentificationrateofseedtypewasabove94%，andtheseedsortingratewasabove97%.Theexperimentalresults showthatthesystemcancompletethesorting taskquicklyandacurately，andcanprovideefective technicalsupportfortheautomation and intellgence of wheat sorting.

Keywords：image recognition;control system; line scan camera;FPGA;wheat sorting;real-timecontrol; delay time

0 引言

流通和消费过程中，不完善粒度的评价是决定小麦品质的重要指标，也是小麦品质评价中常见的技术挑战之一。（剩余9664字）