欢迎来到天天文库
浏览记录
ID:39402525
大小:2.02 MB
页数:56页
时间:2019-07-02
《基于机器视觉的工业机器人分拣系统的研究》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、南京林业大学硕士学位论文基于机器视觉的工业机器人分拣系统的研究姓名:杜荣申请学位级别:硕士专业:机械设计及理论指导教师:焦恩璋20090601AbstractThispaperbasedonMotomanup6robotresearchesandstructuresaclassification-・and-・grabsystemonthebasisofmachinevision.Thesystemismadeupofcamera,imageacquisitioncard,computer,robotandsoftware.Thesystemcontrolstherobotto
2、achievepickingandchoosinginthewaythattherobotgettingimageoftheobjectbythecamem,analyzingtheimagewithsoftware,gettingthecoordinateandclassificationdataoftheobject,thentrackthetargets.Atfirst,thepaperstudiesthecameracalibration,thenimplementstheDirectLinearTransformalgorithmonMatlab,andprovid
3、esaneasysolutiontotransformwordcoordinatesintorobotcoordinates.Then,thepaperstudiesseveraltargetdetectionalgorithmprinciples,concentratesonandimplementsthetargetdetectionbasedonbackgroundsubtractionandgrayimagethresholding.Attheaspectoftargetrecognition,thispaperdesignandtraintheclassifierb
4、yusingthelineardiscriminantfunctionaccordingtotheobject担feamresuchasareaandperimeter.Attheaspectofvisiontrack,thispaperstudiesglobalnearestneighboralgorithmandmulti—hypothesistrackingalgorithm,implementstheGNNalgorithmandkalmanfilterinthec++programminglanguageandf'mallyachievesthetargettrac
5、konconveyor.Finally,thepaperresearchesthePC.basedmotioncontrolofMotomanrobotonthebasisofMotocon32,SOthattheclassification・-and--grabprogrammeCancontrolrobotsgrabtheworkpieceoncontiguousmovingconveyorinrealtime.Intheprocessofresearch,somelimitationofMotocom32hasbeenfound,andthesolutionisprop
6、osedalso.Throughseveralexperimentsandtests,theauthoraccomplishestheresearchandstructureofaprimaryindustrialrobotclassification—and—grabsystem.On-lineexperimentsprovethemsearchresultsofthispaperthatisthissystemasalinkofindustrialautomationhaspracticalsignificanceinapplicationsandreferenceval
7、ueintheory.KeyWords:industrialrobot;conveyor;classification-and—grab;targetrecognition;robotvision学位论文原创性声明本人郑重声明:所呈交的学位论文,是本人在导师的指导下进行的研究工作所取得的成果。尽我所知,除文中已经特别注明引用的内容和致谢的地方外,本论文不包含任何其他个人或集体已经发表或撰写过的研究成果。对本文的研究做出重要贡献的个人和集体,均已在文中以明确方式注明并表示感谢。本人完全意识到本声明的法律结
此文档下载收益归作者所有