基于机器视觉技术的物体表面缺陷检测

何小利1,宋钰2(1.,643000;2.,643000;):本文对物体表面缺陷进行研究和检测.而检测的方法是采用LED环形灯光直接暗视场正面照明方式来提取插座面板划痕图像.具体过程是使用动态阈值分割图像,并采用放射变换区域特征处理及连通区域提取等技术来检测出插座面板划痕.:机器视觉;物体表面;缺陷检测:TP393:A:1009-4970(2011)02-0064-050,.,,.,,.,,.,.1.,LED,,,,(1).11.1:.,.,.,,,,.LED,,,,(2).:2010-07-03:(2008RCYJ04):(1982-),,,,.20112302JournalofLuoyangNormalUniversityFeb.,2011Vo.l30No.221.2.,,,,;,.,.,.,,,,.S={(r,c)R|gminfr,cgmax}(1),ROIRS.gmin=0gmax=2b-1.,gmingmax..,fr,c,gr,c,S={(r,c)R|fr,c-gr,cgdiff}(2)S={(r,c)R|fr,c-gr,c-gdiff}(3),.,.,.,gr,c,.,(2)(3)fr,c.1.3,..,.,:a=R=(r,c)R1=ni-1cei-csi+1(4),a|R|.,4;,4.,..,,.1.4,,,.,,.,,().,.,.1.5,,,4.,,,,.,.,,.652011222.1,,3,LED.32.24,,4,,.,.42.3,,.4.,,4.2,.6:.565;6,.,.,.,.,,.78.,..789978.,.9.6620112,.,.,.4RegionSquares..,.BorderWidthBorderTolerance,.BorderTolerance,,.,.1011RegionSurface..10()112.4,:,.ScratchWidthMax,ScratchWidthMax.,,ScratchGrayDifMfin.12(a),,,.(a),,.(b)(a)4,.(c),.122.5,4.12(b),.,,,,.,6720112.,.,,..12(c).3,HALCON,..,.,,,,.,,,.[1],,,.[J].(),2010,(1).[2],,.BPPID[J].,2008,(2).[3],,,.[J].(),2006,(6).[4][J].,2008,(3).[5].[J].,2006,(2).[6],,.[J].,1993,(4).[7],.[J].,2004,(8).[8],,.BP[J].,1998,(4).[]TheDetectionofSurfaceDefectBasedonMachineVisionTechnologyHEXiaoli1,SONGYu2(1.SchoolofComputerScience,SichuanUniversityofScience&Engineering,Zigong643000,China;2.NetworkAdministration,SichuanUniversityofScience&Engineering,Zigong643000,China)Abstract:Thispaperintroducesasurfacedefectontheresearchandtesting.ThedetectionmethodistouseLEDringlightsdirectlightdarkfieldpositivewaytoextractthesocketpanelscratchimage.Specificprocessistheuseofdynamicthresholdsegmentationimage,andusingradioactivetransformation,theregionalcharacteristicsoftheregionextractionprocessingandconnectivitytechnologiestodetectthesocketpanelscratches.Keywords:machinevision;surface;defectdetection6820112