图像处理考前复习题双语

简答题：1theresearchcontentsofdigitalimageprocessing数字图像处理的研究内容答：1.Imageacquisition采集2.Imagetransform变换3.Imageenhancement增强4.ImageRestoration恢复5.Imagecompression压缩6.ImageSegmentation分割7.ImageRecognitionandUnderstanding识别与理解2数字图像处理方法有哪些，有何不同？答：空间域处理法和变换域处理法（或称频域法）。空间域指图像平面自身，这类方法是以对图像的像素直接处理为基础的。空域处理法主要有下面两大类：邻域处理法点处理法变换域法：数字图像处理的变换域处理方法是首先对图像进行正交变换，得到变换域系数矩阵，然后再施行各种处理，处理后再反变换到空间域，得到处理结果。3数字图像处理中的基本运算有哪些及其应用？点运算PointOperations：对于一幅输入图像，经过点运算将产生一幅输出图像，输出图像中每个像素点的灰度值仅由相应输入像素点的值决定。点运算有时又被称为对比度增强、对比度拉伸或灰度变换。种类：1、线性点运算LinearPointOperations2、非线性点运算NonlinearMonotonicPointOperations。应用1、对比度增强2、光度学标定3、显示标定4、轮廓线5、裁剪代数运算AlgebraicOperations是指对两幅输入图像进行点对点的加、减、乘、除计算而得到输出图像的运算。应用：图像相加，相减，相乘，相除。几何运算GeometricOperations：几何运算可改变图像中各物体之间的空间关系。应用：几何校正GeometricCalibration2、图像配准ImageRegistration4空间域滤波增强技术中的图像锐化与图像平滑有何区别与联系？答：图象锐化是用于增强边缘，导致高频分量增强，会使图象清晰；图象平滑用于去噪，对图象高频分量即图象边缘会有影响。都属于图象增强，改善图象效果。5均值滤波和中值滤波的特点均值滤波器：又称邻域平均滤波器。均值滤波器的概念非常直观。它用滤波掩模确定的邻域内像素的平均灰度值代替图像每个像素点的值。中值滤波器：是一种非线性滤波，它将邻域内所有像素按灰度极进行排序，然后选择中间值作为输出像素值。6简述频率域滤波的一般步骤。},{),(;,),(),(};,{),(1vuGRyxgvuHvuFvuGyxfRvuF7频率域滤波中的5种滤波器比较低通滤波LowpassFilters：保留低频分量，而通过滤波器函数减弱或抑制高频分量的过程。理想低通滤波器(ILPF)巴特沃思低通滤波器指数低通滤波器梯形低通滤波器高通滤波Highpassfiltering：保留高频分量，而通过滤波器函数减弱或抑制高频分量带通和带阻滤波器BandpassandBandstopFilters：带通滤波器允许一定频率范围内的信号通过而阻止其他频率范围内的信号通过，带阻滤波器正好相反。)(yxf，)(vuF，)(yxf，)(vuG，)(yxf，)(yxg，)(yxf，正变换)(vuH，逆变换同态滤波器HomomorphicFilters：是一种在频域中同时将图像亮度范围进行压缩和将图像对比度进行增强的方法。8什么是伪彩色图像处理？常用的伪彩色图像处理方法有哪些伪彩色图像处理是根据特定的准则对灰度值赋以彩色的处理。密度分割法，灰度级－彩色转换法，频域滤波法。9直方图:是灰度级的函数，描述的是图像中具有该灰度级的像素的个数：其横坐标是灰度级，纵坐标是该灰度出现的频率（像素的个数）。性质:1.灰度直方图表征了图像的一维信息2.图像与灰度直方图之间是多对一的映射关系3.子图直方图之和为整图的直方图作用1．作为图像数字化的参数2．图像分割阈值的选择依据3．计算综合光密度填空题：Digitalimageprocessingapproachesfallintotwobroadcategories:（spatialdomainprocessingmethods）and（transformdomain）processingmethods.（spatialdomain）referstotheimageplaneitself,andapproachesinthiscategoryarebasedondirectmanipulationofpixelsinanimage.Therearetwocategories:(neighborhoodprocessing)（PointProcessing）abasicdigitalimageprocessingsystemincludes（imageacquisitionsystem）,（mainframesystem）,（outputequipment）and（storagedevices.）Theessenceofdigitalimageprocessingistoprocessistoprocessa（two-dimensionalmatrix）incomputer.（digitalimageprocessing）assubjectinganumericalrepresentation(atwo-dimensionalrectangulararray)ofanobjecttoaseriesofoperationsinordertoobtainadesiredresult.（digitizing）istheprocessofconvertinganimagefromitsoriginalformintodigitalform.Thedigitizingprocessincludestwobasicsteps:（sampling）and（quantization）.Digitizingthecoordinatevaluesiscalled（sampling）.Digitizingtheamplitudevaluesiscalled（quantization.）（Pixel）istheminimumunitconsistingoftheimage.Eachpixelhasindependentattributes,andeachpixelhastwoattributesatleast:（location）and（gray-levelvalue.)（spatialresolution）isthesmallestdiscernibledetailinanimage.（Gray-levelresolution）similarlyreferstothesmallestdiscerniblechangeingraylevel.ThePrimitiveTypesofDigitalImages（BinaryImage）（GrayLevelImage）（RGBImage）（IndexImage）(gray-levelhistogram)isafunctionshowing,foreachgraylevel,thenumberofpixelsintheimagethathavethatgraylevelThe(median)isthemiddlegraylevelvalueinallthegraylevelvaluesThe(graylevelmode)isthegraylevelvaluethatappearsmostcontinually.The(rangeofgray-levelvalues)isthedifferenceofthebiggestgraylevelvalueandthesmallestone.Colormodel:(RGBColorModel)(HSIColorModel)asamplingapertureallowsthedigitizertoaccesspixelsindividuallywhileignoringtheremainderoftheimage.amechanismforscanningtheimage.Thisprocessconsistsofmovingthesamplingapertureovertheimageinapredeterminedpattern.alightsensorcanmeasurethebrightnessoftheimageateachpixelthroughthesamplingaperture.aquantizerconvertsthecontinuousoutputofthesensorintoanintegervalue.theoutputstoragemedium.Thegray-levelvaluesproducedbythequantizermustbesortedinanappropriateformatforsubsequentcomputerprocessing.Therearetwogeneraldigitizingapproaches,called(scan-indigitizing）and（scan-outdigitizing）Animagedigitizermusthave（alightsources）,（alightsensor）and（ascanningsystem）ImageInputDevices：(Scanner)(Digitalcamera)Devicesthatrecordapermanentimageonpaperorfilearecalled（imagerecordersorhardcopydevices）Theprocessofusingapatternofsoliddotstosimulateshadesofgrayiscalled（ditheringorhalftoning.）TheSpatialTransformation，transformationrelationships：（TranslationOperation）（MagnificationOperation）（RotationOperation）Gray-LevelInterpolation：（NearestNeighborInterpolation）（BilinearInterpolation）（HigherOrderInterpolation）DirectGray-scaleTransformations：（LinearTransformations）（NonlinearTransformation）（OpticalEnhancements）The（histogrammodifying）istoconstructtransformationfunction,changethehistogramoftheimage,inthisway,thechangedimagewillsatisfiedwiththerequirement.（HistogramEqualizations）theoutputimagehasequallymanypixelsateverygraylevel,thehistogramisflat.Themethodusedtogenerateaprocessedimagethathasaspecifiedhistogramiscalled（histogramspecificationorhistogrammatching.）Accordingtospatialfiltersfunctions,thefiltersareclassifiedintotwocategories:(smoothingfilters)and(sharpeningfilters).Theobjectof(smoothingfilters)istoremovethenoiseinth