题目:基于图像的pcb板的断路短路检测技术研究摘要印刷电路板(PCB)是集成各种电子元器件的信息载体，在各个领域得到了广泛的应用。近年来随着印刷电路板生产复杂度和产量的提高，传统PCB缺陷检测方式因接触受限、高成本、低效率等因素，己经逐渐不能满足现代检测需要，因此研究实现一种PCB缺陷的自动检测系统具有很大的现实意义和实用价值。PCB缺陷检测技术中，自动光学检测技术越来越受到重视，其中图像检测法也将成为自动光学检测的主流。本课题在分析国内外对AOI系统中图像识别软件研究成果的基础上，基于图像处理技术、模式识别技术和缺陷检测技术，提出了一种运用参考法和设计规则校验法处理彩色PCB图像进行缺陷检测的方案。该系统主要由光照、CCD摄像机、图像采集卡及计算机图像处理软件组成。其中图像处理软件部分作为本课题的核心,着重研究了其关键功能模块包括图像灰度化、图像滤波、图像锐化、图像识别几个部分算法的选择与设计，并在MATLAB7.0的环境下进行仿真。运用现代成熟的数字图像处理技术，本文实现了PCB缺陷的软件检测方案。在预处理模块中，结合PCB板的特点运用图像预处理手段,首先对彩色图像进行灰度化，其次运用图像滤波,最后通过图像锐化得到高质量的PCB图像。在识别模块中结合电路板的短路、断路缺陷的特征，识别短路和断路故障。提高了生产效率，降低了生产成本。关键词：缺陷检测；图像处理；图像滤波；图像识别ITitleDetectionofShortCircuitandOpenCircuitofPCBBasedonImageAbstractPrintedCircuitBoard(PCB)isaninformationcarrierintegratingvariouselectroniccomponents.Ithasbeenappliedindifferentfieldswidely.WiththehighercomplexityandoutputofPrintedCircuitBoardmanufactureinrecentyears,thetraditionalmethodsofPCBdefectinspectioncannotmeetdemandsofinspectiongraduallybecauseofrestrictedcontact,highcostorlowefficiency.Therefore,thestudyofautomaticdefectinspectionsystemismeaningfulinthePCBproductiontoday.Automaticopticalinspection(AOI)technologyismoreandmoreimportant,inwhichtheimagedetectionwillbecomemainstreamofautomaticopticalinspectioninPCBdefectinspectiontechnology.ImagerecognitionsoftwareonthebasisofresearchresultsinAOIsystemisanalyzedinthesystem.Then,theschemeofPCBcolorimagestreatmentandPCBdefectinspectionwasproposed,inwhichareferencemethodanddesignrulemethodisused.Itisbasedonimageprocessing,patternrecognitionanddefectdetection.Theinspectionsystemiscomposedofseveralparts,includinglight,CCDcameras,capturecardsandcomputerimageprocessingsoftwaresystems.Asthecoreofthisthesis,thedesignandimplementationalgorithmoffunctionalmodulesisthekeyofimageprocessing.Functionalmodulesiscomposedofimagepre-processingasimagegraying,imagefilter,sharpening,andimagerecognition.ThesealgorithmsaresimulatedinMATLAB7.0.ThesoftwaredetectionprogramofPCBdefectisimplementedusingmodernsophisticateddigitalimageprocessingtechnologyinthispaper.Inthepreprocessingmodule,imagepreprocessingmethodsisusedcombiningthecharacteristicsofPCBboard.Inordertogethigh-qualityPCBimage,first,thecolorimageisgrayed,thenthegrayimageisfiltered,andfinally,theimageissharpened.Inthemoduleofimagerecognition,shortcircuitandopencircuitisrecognizedbyanalyzingthecharacteroftwodefects.Thesystemhavehighproductionefficiencyandlowproductioncosts.IIKeywords:Defectinspection;Imageprocessing;ImageFiltering;Imagerecognition目次1绪论.................................................................11.1课题研究的背景.....................................................11.2国内外现状和发展趋势...............................................21.3课题研究的目的和意义...............................................31.4本课题主要研究的内容....
