目录摘要.......................................................3Abstract...................................................4第1章绪论................................................51.1自动识别课题背景·························································51.2机器视觉····································································51.2.1机器视觉的发展概况...........................................................51.2.2机器视觉与图像处理...........................................................71.3图像处理与识别技术·····················································91.4图像处理与识别系统···················································101.4.1关于计算机图像处理系统.................................................101.4.2图像处理与识别系统的构成.............................................111.5斑马线自动识别系统课题研究内容·································12第2章图像处理与识别及图像理解...........................142.1二值图像分析····························································142.1.1阈值运算.............................................................................152.2图像区域分析····························································172.2.1区域与边缘.......................................................................172.3图像处理与识别及图像理解所研究的内容························192.3.1图像处理技术.....................................................................192.3.2图像识别技术.....................................................................202.3.3图像理解.............................................................................212.4图像处理与识别及图像理解的关系·································242.4.1图像处理.............................................................................242.4.2图像理解.............................................................................252.5图像处理工具MATLAB···············································26第三章斑马线自动识别系统主要算法........................283.1边缘检测··································································283.2坎尼(Canny)算子····················································293.3模板匹配算法····························································32第4章基于matlab的斑马线自动识别系统...................354.1系统结构流图····························································354.2系统功能模块分析与实现·············································364.2.1图像分割模块.....................................................................364.2.2模板读取模块.....................................................................404.2.3图像识别模块.....................................................................404.3GUI界面设计及系统测试············································42结论......................................................45致谢......................................................47参考文献..................................................48摘要机器视觉也称图像分析与理解。机器视觉的发展推动智能系统的发展,也拓宽计算机与各种智能机器的研究范围和应用领域。图像处理与识别技术是机器视觉的一个重要组成部分。图像处理与识别技术的发展经历了初创期,发展期,普及期,和实用期4个阶段。20世纪90年代是图像技术的实用化时期,特点就是图像处理的信息量巨大,对处理速度的要求极高。人行道路的斑马线自动识别系统的课题设计,以一幅交通道路识别为例,具体介绍了斑马线自动识别的原理。整个处理过程分为图像预处理、图像边缘提取、图像定位、图像分割、图像识别五大模块,用MATLAB软件编程来实现每一个部分,最后识别出人行道路图像。在研究的同时对其中出现的问题进行了具体分析。关键词:机器视觉图像处理自动识别预处理边缘提取图像定位图像分割图像识别AbstractMachinevisionisalsoImageanalysisandunderstanding.ThedevelopmentofmachinevisionpromotetheprogressofIntelligentsystem,andalsowidentheresearchandapplicationfieldofcomputerandeveryintelligentmachine.Technologyofimageprocessingandrecognitionistheimportantcomponentofmachinevision.Theprogressofimageprocessingandrecognitionhavefourphasesthatisinitialperioddevelopmentperioduniversalperiodandpracticalperiod.Thepracticalperiodofimagetechnologyis1990s20thcentury.Thefeaturesisthattheinformationofimageprocessingtoobig,andsothat,itsprocessingspeedmustbefast.Thecourseoftrafficsignautomatismrecognitionsystem,withonetrafficsignrecognition,theprincipleofthetrafficsignrecognitionisintroducedconcretely.Thisprocesswasdividedintoimagepre-process,imageedgeextraction,imagelocation,imagedivisionandimagerecognition,whichisimplementedseparatedbyusingMATLAB.Thetrafficsignimageisrecognizedatlast.Atthesametime,theproblemsarealsoanalyzed.Andsolvedintheprocess.Keywords:Machinevisionimageprocessingautomatismrecognitionpre-processedgeextractionimagelocationimagedivisionimagerecognition第1章绪论1.1自动识别课题背景人类在征服自然、改造自然和推动社会进步的过程中,面临着自身能力、能量的局限性,因而发明和创造了许多机器来辅助或代替人类完成任务.智能机器,包括智能机器人,是这种机器最理想的形式,也是人类科学研究中所面临的最大挑战之一.智能机器是指这样一种系统,它能模拟人类的功能,能感知外部世界并有效地解决人所能解决问题.人类感知外部世界主要是通过视觉、触觉、听觉和嗅觉等感觉器官,其中约80%的信息是由视觉获取的.因此,对于智能机器来说,赋予机器以人类视觉功能对发展智能机器是及其重要的,也由此形成了一门新的学科—机器视觉(也称机器视觉或图像分析与理解等).机器视觉的发展不仅将大大推动智能系统的发展,也将拓宽计算机与各种智能机器的研究范围和应用领域。1.2机器视觉1.2.1机器视觉的发展概况70年代中期,麻省理工学院(MIT)人工智能(AI)实验室正式开设“机器视觉”(MachineVision)课程,由国际著名学者B.K.P.Horn教授讲授.同时,MITAI实验室吸引了国际上许多知名学者参与机器视觉的理论、算法、系统设计的研究,DavidMarr教授就是其中的一位.他于1973年应邀在MITAI实验室领导一个以博士生为主体的研究小组,1977年提出了不同于"积木世界"分析方法的计算视觉理论(computationalvision),该理论在80年代成为机器视觉研究领域中的一个十分重要的理论框架.可以说,对机器视觉的全球性研究热潮是从20世纪80年代开始的,到了80年代中期,机器视觉获得了蓬勃发展,新概念、新方法、新理论不断涌现,比如,基于感知特征群的物体识别理论框架,主动视觉理论框架,视觉集成理论框架等.到目前为止,机器视觉仍然是一个非常活跃的研究领域.许多会议论文集都反应了该领域的最新进展,比如,InternationalConferenceonComputerVisionandPatternRecognition(CVPR);InternationalConferenceonComputerVision(ICCV);InternationalConferenceonPatternRecognition(ICPR);Internationa