Roboticvisiontechniqueisallimportantresearchfieldonthemachinevision,andifsanewsubjectbasedonthedevelopmentofthecomputervisiontheoriesandtheimageprocessingtechniques.Ithasbrokenthetraditionalmechanicalarm谢nIdemonstrationteachingandoff-lineprogrammingcontrolmode,anddon'tneedtobetothemovementforprogramming,greatlyimprovingthemechanicalarmflexibility,canbewidelyusedinservicerobotandindustrialrobots,etc.Inthispaper,basedontheChinesemassagerobotsystemswith4degreeoffreedom(ooF),therecognitionandtrackingOnthesingletargetobjectandmulti=targetobjectsarestudiedbyusingthemethodwitIlposition-basedvisualSeI'VO,andsomefindingsandconclusionshavebeenachieved.Thisresearchcontentsincludethefollowingrespects.
Thefirst,thestructureofChinesemassagerobotisanalyzed.WimtheD-Hmethod,thetransformationmatrixisobtainedandthemodeloftheChinesemassagerobot’Skinematicsissetup.Tosolvethekinematicsequationis.Thepositivesolutionofthekinematicequationsandinversesolutionaresolved.ThenwefindtherelationshipbetweenpositionandjointturnIll
山东建筑大学硕士学位论文
variables.
Thesecond,bytheanalysisoftheprincipleofcameraimagingandunderstanding,combiningthecurrentc组Jlleracalibrationmethod,aquickandconvenientmonocularvisionmeasurementofthecameracalibration
processofrobotmethodisproposed.Thismethodmethodsimplifiestheandcameracalibration.Feasibilityofthisisverifiedbyexperiments.
Thethird,for
papermakeaidentificationandlocationoftheacupuncturepointsofthebodyback,thisandcontrasttheiradvantagesdetailedanalysisanddisadvantagestothetraditionalimagematchingalgorithm.Basedofthese,thispapertakesactualartificialsymbolforthetarget,Basedonfastrelatedgrayimagematchingnormalizedalgorithmacupoints
onpositionednew
positionedmethodsandbasedtheSWTcharacteristicsofthematchingexperimentalanalysisusedtheactualacupointsnewmethoda托proposed.Theimage.Theresultsshowedthatbothofthemhavegoodstability.
The
isfour,inthispaper,an∞ⅣlegonstancedetectionmethodbasedOntheHoughalgorithmproposedthebasisofdeepanalysisofthecharacteristicsandthelegsHoughtransform
takeprinciple.nIemethod
realizetheartificialadvantageofthealgorithmofHoughlinedetectionefficiencyandlegflaglinefeaturedetection.nleresultsshowedthatthemethodcouldeffectivelyextractthecharacteristicsofartificialsignsandachievethedesiredresults.
Thefive,CompletionoftheoverallstructuraldesignofChinesemassage
vision.Introducedfromthesystemsoft,rarerobotbasedOilarchitecture,programflow,andtherecognitionalgorithm’Simplementfunctions.Completedthelayoutofpowercordandcontrolsignallinesincontrolcabinetandensuredthesafeandstableoperationoftheentiremoiloncontrolsystem.