基于Hu矩和Zernike矩的图像目标识别算法设计

基于Hu矩和Zernike矩的图像目标识别算法设计院系自动化学院专业测控技术与仪器班级5407101学号200504071002姓名郭晓宇指导教师李忠海负责教师李忠海沈阳航空工业学院2009年6月沈阳航空工业学院毕业设计（论文）--I摘要形状识别在计算机视觉中具有十分重要的意义，利用矩特征进行形状识别是一种重要的方法。近几年用正交矩进行图像分析，图像处理以及图像识别的研究成果很多。这表明不变矩理论及其在图像信息处理与识别的应用技术具有很好的发展前景和商机。理论上矩不变量在图像平移、伸缩、旋转时均保持不变，这为图像识别算法中目标矩特征的选择提供了一定的依据。不变矩是一种高度浓缩的图像特征，具有平移、尺度、旋转等不变性。1961年，M.K.Hu首先提出了7个不变矩用于图像描述。后来人们进行了多方面的研究，发现正交矩具有绝对的独立性，没有信息冗余现象，抽样性能好，抗噪声能力强，适合于图像识别。本文主要完成对航拍图像中的飞行目标进行识别的系统设计工作。主要阐述了将Hu矩和Zernike矩作为目标在旋转、平移和缩放等变形的情况下不变的目标特征的可行性，并比较他们在具体应用中的特点。设计了如下算法流程：首先，对图像进行预处理，将彩色图像转化为灰度图像，再进行二值化处理；然后利用Hu不变矩和Zernike矩提取不变矩特征；最后，采用街区距离法，计算出目标区域值，实现了对航拍图像中飞行目标的自动识别。实验结果表明，不变矩在图像中的目标识别具有良好的稳定性，Hu矩和Zernike矩对目标的识别具有很好的作用。关键词：Hu矩；Zernike矩；矩不变量；目标识别沈阳航空工业学院毕业设计（论文）--IIAbstractShaperecognitionisaveryimportantproblemincomputervision.Recognitingfyingtargetswithmomentfeaturesisanimpotantmethodforshapeidentification.Inrecentyears，manyresultshavebeenresearchedaboutimageanalysisandpatternrecognitionwithorthogonalmoments.Therefore，thetheoryofinvariantmomentsandtheirapplicationtoimageanalysisandpatternrecognitionhaveagoodfuture.Invariantmomentsareindependentoftranslation，scaleandrotationintheory.Theresultsofsuchcomparisoncanprovidesomebaseswhichwouldbearpracticabilityfortheselectionofmomentfeatureinimagerecognition.Invariantmomentsarehighlyconcentratedimagefeaturesthatareshiftinvariant，rotationinvariantandscaleinvariants.M.K.Hufirstintroducedsevenmomentinvariantsin1961，basedonmethodsofalgebraicinvariants.Laterstudiesindicatedthattheorthogonalmomentshavethebestoverallperformanceintermsofnoisesensitivity，informationredundancy，andcapabilityofimagedescription.Inthispaper，thecompletionofaerialflightinthetargetidentification.IthighlightsamethedofusingHuandZernikemomentastargetfeature，comparingthemmomentfeatures.Thefollowingalgorithmdesignprocess:Firstly，makepretreatmentfortheimage，thatis，transformthemulticolorimagetograyimageandtodealwithbinary.ThencomputeinvariantmomentcharactersbyHuinvariantmomentandzernikemoment.Atlast，thisarticlepresentsasquarespacemethodthatcanbeusedtocalculatetheareavalueoftheobjects.Achievedontheaerialimagestoautomaticidentifytheobjectivesoftheflight.Experimentresultsshowthatinvariantmomenthaveperfitstabilityinimagetargetrecognition.Humomentandzernikemomentplayagoodroleinidentifyingobjects.Keywords:Humoment；Zernikemoment；invariantmoment；targetrecognition沈阳航空工业学院毕业设计（论文）--III目录第1章绪论.....................................................................................................................11.1课题背景.....................................................................................................................11.1.1图像目标识别的简单介绍...................................................................................21.1.2识别目标技术的发展概况...................................................................................31.2课题任务及要求.........................................................................................................51.3课题内容及安排.........................................................................................................5第2章目标识别算法中不变矩基本原理.........................................................................62.1基于Hu矩的目标识别..............................................................................................62.1.1不变矩定义...........................................................................................................62.1.2Hu矩原理.............................................................................................................62.1.3Hu矩计算方法.....................................................................................................72.2基于Zernike矩的目标识别.......................................................................................92.2.1Zernike矩多项式.................................................................................................92.2.2Zernike矩定义...................................................................................................102.2.3Zernike矩计算方法...........................................................................................10第3章图像目标识别算法的实现...................................................................................133.1算法方案的整体流程图...........................................................................................133.2图像的获取与处理...................................................................................................143.2.1航拍图像的获取.................................................................................................143.2.2航拍图像的预处理.............................................................................................143.3Hu矩算法的实现.....................................................................................................183.3.1Hu矩算法的流程设计.......................................................................................183.3.2提取Hu矩的特征向量......................................................................................193.4基于Hu矩不变性的识别.......................................................................................213.5Zernike矩算