基于视觉技术的苹果分拣系统设计

0，，，。，。、、，，，，，、、。，。，、、、，，。，，、，[1]。，，，。，、、、、、。1Matlab，，：1）。，。2）。，，。3）。RGBHIS，。4）。，。，，，。。5），。，，。：1）。CCD、。2）。、、、王鑫1，赵莹2，杨简1(1.，，130118；2.，，133613)：2013129：20131225：，，1982，，，；。E-mail:[e-mail]：，。，，，，、，。94.2%，，。：；；；：S226.5：A：2095-5553(2014)05-0169-04，，.[J].,2014,35(5):169~172WangXin,ZhaoYing,YangJian.Designoffruitssortingsystembasedonvisiontechnology[J].JournalofChineseAgriculturalMechanization,2014,35(5):169~172DOI:10.13733/j.jcam.issn.2095-5553.2014.05.039JournalofChineseAgriculturalMechanization35520149Vol.35No.5Sep.2014、。3）。。4）。RGB。5）。，，，，。6）。。22.1。，。1。，CDD、、。：，。，。2.2、、、、、。2.2.1，，、，。rgb=imread（‘e:/77.jpg’）;figure,subplot（131）,imshow（rgb）,title(‘’）;gray=rgb2gray(rgb);subplot(132),imshow(gray),title(‘’）;threshold=graythresh(gray);%bw=im2bw(gray,threshold);subplot(133),imshow(bw),title(‘’）;j1=medfilt2(gray);figure,subplot(121）,imshow(j1);title(‘’）j2=medfilt2(bw);subplot(122),imshow(j2);ti-tle('');%RGB2，3。23，，，。。2.2.2，。，。，，，，[2]。4。%%%2.2.3。，、、。，2RGBFig.2GrayscaleandtwovalueimageofRGBmap3Fig.3Filtereffectforthegrayscaleandtwovalueimage4Fig.4Grayscaleadjustmentimageoftheapplewithscar1Fig.1Imageacquisitiondevice1.CCD2.3.4.2014170，sobel、prewitt。5。3、、。，，[3]。()、、。3.1，、、。。，。，，[4]。：xo=12x,y∈boundary%%矣xydx-12x2d矣矣y12x,y∈boundary%%矣（ydx-xdy）yo=12x,y∈boundary%%矣12y2dx-xyd矣矣x12x,y∈boundary%%矣（ydx-xdy），[5]。L=bwlabel(BW1,8);%stats=regionprops(L,‘all’);centroids=cat(1,stats.Centroid);imshow(BW1);title(‘’)holdon%plot(centroids(:,1),centroids(:,2),‘g*’)%g*-*holdoffarea=[stats.Area];R=sqrt(area/pi)R=27.41406。3.2，[6]，，RGB。RGBHIS，H。H。3.3，，，，。：→→→→→。7。3.4，、、、、。1。，。5Fig.5Edgedetectionoftheapplewithscar6Fig.6Centroidoftheapplemap7Fig.7Defectextractionoftheapplewithscar/mm/cm28025°0.07545°0.17065°0.56580°1.01Tab.1Standardqualityofapple:5171、、。，。，max-min。4，，30，120，，，，，2。5，，、、、，，。，，94.2%。，，。[1],,.[J].,2010,(6):56~57.QinYonghui,WangWei,etal.Designofroller-typeclassifier[J].AgriculturalScience&TechnologyandEquipment,2010,(6):56~57.[2],.[J].,2012,(2):141~144.FengBin,WangMaohua.Computervisionclassificationoffruitbasedonfractalcolor[J].TransactionsoftheChineseSocietyofAgriculturalEngineering,2012,(2):141~144.[3],.[J].,2002,18(4):163~166.HuangYonglin,YingYibin.Controllerforfruitsynchronoustrackingandauto-classificationusedinreal-timefruitgradingsystem[J].TransactionsoftheChineseSocietyofAgriculturalEngineering,2002,18(4):163~166.[4]LeemansV,MageinH,etal.On-linefruitgradingaccordingtotheirex-ternalqualityusingmachinevision[J].BiosystemsEngineering,2002,83(4):397~404.[5]BennedsenB.S,PetersonD.L.Indentificationofapplestemandcalyxus-ingunsupervisedfeatureextraction[J].TransactionsoftheASAE,2004,47(3):889~894.[6]DiazR,Gil.L,etal.Comparisonofthreealgorithmsintheclassificationoftableolivesbymeansofcomputervision[J].Journaloffoodengi-neering,2004,(61):101~107.///%3028293.33029196.73028293.33022893.32Tab.2ClassificationresultsofapplesDesignoffruitssortingsystembasedonvisiontechnologyWangXin1,ZhaoYing2,YangJian1(1.InformationtechnologyteachingandManagementCenter,JilinAgriculturalUniversity,Changchun,130118,China;2.ChangbaiMountainInstituteofScience,Yanbian,133613,China)Abstract:On-linedetectionisacriticalpartofautomaticproduction,andvisualimageinformationprocessingtechniquehasbeengraduallyappliedinthefieldoffruitqualitydetection.Surfacedefectdetectionisadifficultproblemforfruitautomaticgrading.Thefeatureextractionmethodwasusedwhichfirstgetstheextractionofdefectivepartsandfillstheareatodeterminethedefectsize,thencombinesmaximumtransversediameterandcolortonewithdefectsizetorealizeapplesortingmechanization.Thismethodhas94.2%precisionrateandbetterstability,couldprovideareliablemethodtoacceleratefruiton-linedetectionforagricultureautomation.Keywords:qualitydetection;computervision;imageprocess;appleclassification2014172