基于DSP的啤酒瓶视觉检测系统设计与开发

《基于DSP的啤酒瓶视觉检测系统设计与开发》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、浙江大学硕士学位论文基于DSP的啤酒瓶视觉检测系统设计与开发姓名：王程申请学位级别：硕士专业：检测技术与自动化装置指导教师：周泽魁20050201浙江大学硕士毕业论文ABSTRACTChinaisthetopbreweroftheworld,andbeerindustryplaysanimportantroleinthedevelopmentofnationaleconomy.However,thetechnologylevelofourbrewhouses'equipmentslagsbehindabroad,andgreatdisparityexitsinbotl

2、ingprocessofbeerproduction.Beerbotlemustbeinspectedbeforebeingbotled.Traditionalinspectionmodewhichreliesonhumaneyescannotguaranteetheaccuracyduetothehighspeedofassemblyline.Thesystembasedonmachinevisionhastheadvantageofhighspeed,highprecisionandhighreliability.Itcaninspectobjectwithout

3、touching.Adoptingthissysteminbeerbottleinspectioncanovercometheweaknessofhumaneyesandimproveautomationlevelandproductivitygreatly.Toinspectbeerbotleonassemblyline,theinspectionsystemmustbecapapleofprocessingvisionimageinreal-time.DigitalSignalProcessorisabletomatchthisdamandwell.Themajo

4、rcontributionsinthisdissertationareasfollows:Anemptybotlevisioninspectionsystemisdesigned,focusingonthecharacteristicsofbeerbotleinspectionandtheactualrequirementsofproduction.CCDtechnology,opticstechnology,controltechnologyandrealtimeimageprocessingtechnologyareemployedinthedesigntogua

5、ranteeaccuracyandinspectionspeed.ArealtimeimageprocessingsystembasedonDSPhasbeendeveloped.ThesystemcanbeusedinmanyvisionsystemswiththestructureofDSP+CPLD.InPCBdesign,thelayouttechniqueofhighfrequencyisadoptedtopreventthepotentialEMCproblems.ThesoftwaredesignofDSPisaccomplished,focusingo

6、nthedemandofinspectionandthecharacteristicofDSP.Thestorageandtransferofimagedataarediscussed,whichmadethesystemcapableofprocessingmassdatarealtimewithlimitedhardwareresource.FlexiblesettingofinterruptandDMAisadoptedinthesoftwaredesign,whichimprovestheefficiencyofdatatransferandperforman

7、ceofDSP.Keywords:EmptyBotleInspection,MachineVision,ImageCollection,RealtimeImageProcessing,DSP浙江大学硕士毕业论文第1章绪论1.1机器视觉的概况与发展1.1.1机器视觉的定义美国制造工程师协会(SMESocietyofManufacturingEngineers)机器视觉分会和美国机器人工业协会(RIARoboticIndustriesAssociation)的自动化视觉分会对机器视觉下的定义为:“机器视觉是通过光学的装置和非接触的传感器自动地接收和处理一