数字图像打印要点Word格式文档下载.docx
《数字图像打印要点Word格式文档下载.docx》由会员分享，可在线阅读，更多相关《数字图像打印要点Word格式文档下载.docx（19页珍藏版）》请在冰点文库上搜索。
Smoothingfilter/convolution平滑滤波器/卷积
smoothing
filter
are
used
for
blurring
and
noise
reduction平滑滤波器用于模糊处理和降低噪声/卷积
Imageenhancement/imagerestoration图像增强和图像恢复
空间域滤波Spatialdomainfiltering：
频率域滤波Frequencydomainfiltering：
Frequencydomainfilteringwithavariablefrequencyforthesignalfiltering以频率作为变量对信号进行滤波
空间分辨率：
spatialresolutionisameasureofthesmallestdiscernibledetailinanimage.图像中可辨别的最小细节的度量
灰度分辨率：
Intensityresolutionreferstothesmallestdiscerniblechangeinintensitylevel.灰度分辨率是指在灰度级中可分辨的最小变化
取样sampling:
Digitizingthecoordinatevaluesiscalledsampling.对坐标值进行数字化
量化quantization：
Digitizingtheamplitudevaluesiscalledquantization.对幅值数字化
图像压缩：
Imagecompression,theartandscienceofreducingtheamountofdatarequiredtorepresentanimage.图像压缩是一种减少描绘一幅图像所需数据量的技术和科学.
（2）问答题
1.Citeoneexampleofdigitalimageprocessing
Answer:
Inthedomainofmedicalimageprocessingwemayneedtoinspectacertainclassofimagesgeneratedbyanelectronmicroscopetoeliminatebright,isolateddotsthatarenointerest.
2.Citeoneexampleofspatialoperation举一个空间操作的例子
3.Citeoneexampleoffrequencydomainoperationfromthefollowingprocessingresult,makeageneralcommentaboutidealhighpassfilter（figureB）andGaussianhighpassfilter（figureD）
A.Originalimage
B.idealhighpassfilter
Incontrasttotheideallowpassfilter,itistoletallthesignalsabovethecutofffrequencyfcwithoutloss,andtomakeallthesignalsbelowthecutofffrequencyofFCwithoutlossof.
C.theresultofidealhighpassfilter
D.Gaussianhighpassfilter
Highpassfilter,alsoknownas"
lowresistancefilter"
itisaninhibitoryspectrumofthelowfrequencysignalandretainhighfrequencysignalmodel（ordevice）.Highpassfiltercanmakethehighfrequencycomponents,whilethehigh-frequencypartofthefrequencyintheimageofthesharpchangeinthegrayarea,whichisoftentheedgeoftheobject.Sohighpassfiltercanmaketheimagegetsharpeningprocessing
E.TheresultofGaussianfilter
3.Theoriginalimage,theideallowpassfilterandGaussianlowpassfilterareshownbelowBndC.DandEaretheresultoftheeitherfilterBorC
A.Drawlinestoconnectthefilterwiththeirresult
B.Explainthedifferenceofthetwofilters
Duetoexcessivecharacteristicsoftheideallow-passfiltertoofastJun,itwillproducearingingphenomenon.OvercharacteristicsofGaussfilterisveryflat,soitisnotringing
4.Whatistheresultwhenapplyinganaveragingmaskwiththesize1X1?
Nochange
5.StatetheconceptoftheNyquistsamplingtheoremfromthefigurebelovy
Thelawofsamplingprocessshouldbefollowed,alsocalledthesamplingtheoremandthesamplingtheorem.Thesamplingtheoremshowstherelationshipbetweenthesamplingfrequencyandthesignalspectrum,anditisthebasicbasisofthecontinuoussignaldiscretization.Inanalog/digitalsignalconversionprocess,whenthesamplingfrequencyfs.maxgreaterthan2timesthehighestfrequencypresentinthesignalFmaxfs.max>
2fmax,samplingdigitalsignalcompletelyretainedtheinformationintheoriginalsignal,thegeneralpracticalapplicationassurancesamplingfrequencyis5~10timeshigherthanthatofthesignalofthehighfrequency;
samplingtheorem,alsoknownastheNyquisttheorem
6.Ameanfilterisalinearfilterbutamedianfilterisnot,why?
Thebasicprincipleoflinearfilteringistoreplacetheoriginalimagewiththemeanvalueofeachpixel,butmedianfilterreplacetheoriginalimagewiththemedianvalueofeachpixel.Thevalueofmeanandmedianisdifferent.
7.Fundamental
Steps
in
images
Digital
image
Processing
数字图像图像处理的基本步骤
acquisition—>
enhancement—>
restoration—>
Color
processing—>
wavelets—>
compression（压缩）—>
morphological
processing（形态学理）—>
segmentation（分割）—>
representation
description（表示与描述）—>
recognition（识别）
8.WiththechromaticitydiagrambellowgiveabriefdescriptiontotheRGBcolormodel.Andthesethreecolorsenoughtocomposeallvisiblecolors?
Answer：
ImagesrepresentedintheRGBcolormodelconsistofthreecomponentimages,oneforeachprimarycolor.
Thesethreecolorsenoughtocomposeallvisiblecolors
（3）算法题
1.ThefollowingmatrixAisa3*3imageandBis3*3Laplacianmask,whatwillbetheresultingimage?
（Notethattheelementsbeyondtheborderremainunchanged）
2.DevelopanalgorithmtoobtaintheprocessingresultBfromoriginalimageA
3.Developanalgorithmwhichcomputesthepseudocolorimageprocessingbymeansoffouriertramsform
Thestepsoftheprocessareasfollow:
（1）Multiplytheinputimagef（x,y）by（-1）x+ytocenterthetransform;
（2）ComputetheDFToftheimagefrom
（1）togetpowerspectrumF（u,v）ofFouriertransform.
（3）Multiplybyafilterfunctionh（u,v）.
（4）ComputetheinverseDFToftheresultin（3）.
（5）Obtaintherealpartoftheresultin（4）.
（6）Multiplytheresultin（5）by（-1）x+y
4.Developanalgorithmtogenerateapproximationimageseriesshowninthefollowingfigureb**meansofdownsampling
5.DevelopanalgorithmwhichimplementsfrequencydomainfilteringbymeansofFouriertransform.
（1）将输入图像f（x，y）的（-1）x+y为中心的变换；
计算图像的DFT从
（1）得到的功率谱f（u，v）的傅里叶变换。
Multiplybyafilterfunctionh（u,v）
乘以一个滤波器函数h（u,v）.
ComputetheinverseDFToftheresultin（3）.
计算（3）中的结果DFT的逆
Obtaintherealpartoftheresultin（4）.
获得（4）结果中的实部
Multiplytheresultin（5）by（-1）x+y
（-1）x+y乘以（5）中的结果.
5.Losslessapproaches—HoffmanCoding无损方法-霍夫曼编码
步骤：
（1）createofsourcereductionsbyorderingthesymbolsunderconsiderationandcombiningthelowestprobabilitysymbolsintoasinglesymbolsthatreplacestheminthenextsourcereduction.
（2）Codeeachreducedsource,startingwiththesmallestsourceandworkingbacktotheoriginalsource.
（4）编程题
1）Therearetwosatellitephotosofnightasblew.WriteaprogramwithMATLABtotellwhichisbrighter
代码：
A=imread（’1.jgp’）;
B=imread（‘2.jpg’）;
[m,n]=size（A）;
fori=1:
m
forj=1:
n
sum1=sum1+A[I,j];
end
avg1=sum1/m*n;
[r,c]=size（B）;
sum2=sum2+B[I,j];
avg2=sum2/m*n;
2）An8*8imagef（i,i）hasgraylevelsgivenbythefollowingequation:
f（i,i）=|i-j|,i,j=0,1….,7
Writeaprogramtofindtheoutputimageobtainedbyapplyinga3*3medianfilterontheimagef（i,j）;
notethattheborderpixelsremainunchanged.
Ansewr:
function[r]=avgfilter（gray,n）
a（1:
n,1:
n）=1;
[row,col]=size（gray）;
gray1=double（gray）;
gray2=gray1;
row-n+1
forj=1:
col-n+1
c=gray1（i:
i+（n-1）,j:
j+（n-1））.*a;
s=sum（sum（c））;
gray2（i+（n-1）/2,j+（n-1）/2）=s/（n*n）;
end
end
r=uint8（gray2）;
>
avg3=avgfilter（noise,3）;
avg5=avgfilter（noise,5）;
avg7=avgfilter（noise,7）;
subplot（221）;
imshow（noise）;
title（'
原噪声图'
）;
subplot（222）;
imshow（avg3）;
3*3均值滤波图'
subplot（223）;
imshow（avg5）;
5*5均值滤波图'
subplot（224）;
imshow（avg7）;
7*7均值滤波图'
1．DesignanadaptivelocalnoisereductionfilterandapplyittoanimagewithGaussiannoise.Comparetheperformanceoftheadaptivelocalnoisereductionfilterwitharithmeticmeanandgeometricmeanfilter.
clear
closeall;
rt=imread（'
E:
\数字图像处理\yy.bmp'
gray=rgb2gray（rt）;
subplot（2,3,1）;
imshow（rt）;
原图像'
）;
subplot（2,3,2）;
imshow（gray）;
原灰度图像'
rtg=im2double（gray）;
rtg=imnoise（rtg,'
gaussian'
0,0.005）%加入均值为0，方差为0.005的高斯噪声
subplot（2,3,3）;
imshow（rtg）;
高噪点处理后的图像'
[a,b]=size（rtg）;
n=3;
smax=7;
nrt=zeros（a+（smax-1）,b+（smax-1））;
fori=（（smax-1）/2+1）:
（a+（smax-1）/2）
forj=（（smax-1）/2+1）:
（b+（smax-1）/2）
nrt（i,j）=rtg（i-（smax-1）/2,j-（smax-1）/2）;
end
figure;
imshow（nrt）;
扩充后的图像'
nrt2=zeros（a,b）;
fori=n+1:
a+n
forj=n+1:
b+n
form1=3:
2
m2=（m1-1）/2;
c=nrt2（i-m2:
i+m2,j-m2:
j+m2）;
%使用7*7的滤波器
Zmed=median（median（c））;
Zmin=min（min（c））;
Zmax=max（max（c））;
A1=Zmed-Zmin;
A2=Zmed-Zmax;
if（A1>
0&
&
A2<
0）
B1=nrt2（i,j）-Zmin;
B2=nrt2（i,j）-Zmax;
if（B1>
B2<
0）
nrt2（i,j）=nrt2（i,j）;
else
nrt2（i,j）=Zmed;
continue;
nrt3=im2uint8（nrt2）;
imshow（nrt3）;
自适应中值滤波图'
2.ImplementWienerfilterwith“wiener2”functionofMatLabtoanimagewithGaussiannoiseandcomparetheperformancewithadaptivelocalnoisereductionfilter.
代码如下：
I=imread（'
J=rgb2gray（I）;
K=imnoise（J,'
0,0.005）;
L=wiener2（K,[55]
升级会员 