数字图像处理冈萨雷斯英文ChapterEng 数字图像基础.pptx

《数字图像处理冈萨雷斯英文ChapterEng 数字图像基础.pptx》由会员分享，可在线阅读，更多相关《数字图像处理冈萨雷斯英文ChapterEng 数字图像基础.pptx（56页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、What is Digital Image Processing?Processing of a multidimensional pictures by a digital computerWhy we need Digital Image Processing?1.To record and store images2.To enhance images using mathematics3.To analysis images4.To synthesize images5.To create computer vision systems第1页/共56页Digital Image Dig

2、ital image=a multidimensionalarray of numbers(such as intensity image)or vectors(such as color image)Each component in the imagecalled pixel associates withthe pixel value(a single number in the case of intensity images or a vector in the case of color images).第2页/共56页Visual Perception:Human Eye(Pic

3、ture from Microsoft Encarta 2000)第3页/共56页Cross Section of the Human Eye(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第4页/共56页1.The lens contains 60-70%water,6%of fat.2.The iris diaphragm controls amount of light that enters the eye.3.Light receptors in the re

4、tina-About 6-7 millions cones for bright light vision called photopic-Density of cones is about 150,000 elements/mm2.-Cones involve in color vision.-Cones are concentrated in fovea about 1.5x1.5 mm2.-About 75-150 millions rods for dim light vision called scotopic-Rods are sensitive to low level of l

5、ight and are not involved color vision.4.Blind spot is the region of emergence of the optic nerve from the eye.Visual Perception:Human Eye(cont.)第5页/共56页Range of Relative Brightness SensationSimutaneous range is smaller thanTotal adaptation range(Images from Rafael C.Gonzalez and Richard E.Wood,Digi

6、tal Image Processing,2nd Edition.第6页/共56页Distribution of Rods and Cones in the Retina(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第7页/共56页Image Formation in the Human Eye(Picture from Microsoft Encarta 2000)(Images from Rafael C.Gonzalez and Richard E.Wood,D

7、igital Image Processing,2nd Edition.第8页/共56页PositionIntensityBrightness Adaptation of Human Eye:Mach Band Effect第9页/共56页Mach Band EffectIntensities of surrounding points effect perceived brightness at each point.In this image,edges between bars appear brighter on the right side and darker on the lef

8、t side.(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第10页/共56页 In area A,brightness perceived is darker while in area B isbrighter.This phenomenon is called Mach Band Effect.PositionIntensityABMach Band Effect(Cont)第11页/共56页 Simultaneous contrast.All small sq

9、uares have exactly the same intensitybut they appear progressively darker as background becomes lighter.Brightness Adaptation of Human Eye:Simultaneous Contrast第12页/共56页Simultaneous Contrast(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第13页/共56页Optical illusi

10、on(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第14页/共56页Visible Spectrum(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第15页/共56页Image SensorsSingle sensorLine sensorArray sensor(Images from Rafael C.Gonzalez and Richar

11、d E.Wood,Digital Image Processing,2nd Edition.第16页/共56页Image Sensors:Single Sensor(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第17页/共56页Image Sensors:Line SensorFingerprint sweep sensorComputerized Axial Tomography(Images from Rafael C.Gonzalez and Richard E

12、.Wood,Digital Image Processing,2nd Edition.第18页/共56页CCD KAF-3200E from Kodak.(2184 x 1472 pixels,Pixel size 6.8 microns2)Charge-Coupled Device(CCD)w Used for convert a continuous image into a digital imagew Contains an array of light sensorsw Converts photon into electric chargesaccumulated in each

13、sensor unitImage Sensors:Array Sensor第19页/共56页Horizontal Transportation RegisterOutput GateAmplifierVertical Transport RegisterGateVertical Transport RegisterGateVertical Transport RegisterGatePhotositesOutputImage Sensor:Inside Charge-Coupled Device第20页/共56页Image Sensor:How CCD worksabcghidefabcghi

14、defabcghidefVertical shiftHorizontal shiftImage pixelHorizontal transportregisterOutput第21页/共56页Image“After snow storm”Fundamentals of Digital Imagesf(x,y)xyw An image:a multidimensional function of spatial coordinates.w Spatial coordinate:(x,y)for 2D case such as photograph,(x,y,z)for 3D case such

15、as CT scan images (x,y,t)for movies w The function f may represent intensity(for monochrome images)or color(for color images)or other associated values.Origin第22页/共56页Digital ImagesDigital image:an image that has been discretized both inSpatial coordinates and associated value.w Consist of 2 sets:(1

16、)a point set and(2)a value setw Can be represented in the formI=(x,a(x):x X,a(x)F where X and F are a point set and value set,respectively.w An element of the image,(x,a(x)is called a pixel where-x is called the pixel location and-a(x)is the pixel value at the location x第23页/共56页Conventional Coordin

17、ate for Image Representation(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第24页/共56页Digital Image Types:Intensity ImageIntensity image or monochrome image each pixel corresponds to light intensitynormally represented in gray scale(gray level).Gray scale values

18、第25页/共56页Digital Image Types:RGB ImageColor image or RGB image:each pixel contains a vectorrepresenting red,green andblue components.RGB components第26页/共56页Image Types:Binary ImageBinary image or black and white imageEach pixel contains one bit:1 represent white0 represents blackBinary data第27页/共56页

19、Image Types:Index ImageIndex imageEach pixel contains index numberpointing to a color in a color tableIndex valueIndex No.RedcomponentGreencomponentBluecomponent10.10.50.321.00.00.030.01.00.040.50.50.550.20.80.9Color Table第28页/共56页Digital Image Acquisition Process(Images from Rafael C.Gonzalez and R

20、ichard E.Wood,Digital Image Processing,2nd Edition.第29页/共56页Generating a Digital Image(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第30页/共56页Image Sampling and QuantizationImage sampling:discretize an image in the spatial domainSpatial resolution/image resolu

21、tion:pixel size or number of pixels(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第31页/共56页How to choose the spatial resolution=Sampling locationsOriginal imageSampled imageUnder sampling,we lost some image details!Spatial resolution第32页/共56页How to choose the

22、spatial resolution:Nyquist RateOriginal image=Sampling locationsMinimumPeriodSpatial resolution(sampling rate)Sampled imageNo detail is lost!Nyquist Rate:Spatial resolution must be less or equalhalf of the minimum period of the imageor sampling frequency must be greater orEqual twice of the maximum

23、frequency.2mm1mm第33页/共56页Sampling rate:5 samples/secAliased Frequency Two different frequencies but the same results!第34页/共56页Effect of Spatial Resolution256x256 pixels64x64 pixels128x128 pixels32x32 pixels第35页/共56页Effect of Spatial Resolution(Images from Rafael C.Gonzalez and Richard E.Wood,Digital

24、 Image Processing,2nd Edition.第36页/共56页Moire Pattern Effect:Special Case of SamplingMoire patterns occur when frequencies of two superimposed periodic patterns are close to each other.(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第37页/共56页Effect of Spatial Re

25、solution(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第38页/共56页Can we increase spatial resolution by interpolation?Down sampling is an irreversible process.(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第39页/共56页Image Q

26、uantizationImage quantization:discretize continuous pixel values into discrete numbersColor resolution/color depth/levels:-No.of colors or gray levels or-No.of bits representing each pixel value-No.of colors or gray levels Nc is given bywhere b=no.of bits第40页/共56页Quantization functionLight intensity

27、Quantization level012Nc-1Nc-2DarkestBrightest第41页/共56页Effect of Quantization Levels256 levels128 levels32 levels64 levels第42页/共56页Effect of Quantization Levels(cont.)16 levels8 levels2 levels4 levels In this image,it is easy to seefalse contour.第43页/共56页How to select the suitable size and pixel dept

28、h of images Low detail imageMedium detail imageHigh detail imageLena imageCameraman imageTo satisfy human mind1.For images of the same size,the low detail image may need more pixel depth.2.As an image size increase,fewer gray levels may be needed.The word“suitable”is subjective:depending on“subject”

29、.(Images from Rafael C.Gonzalez and Richard E.Wood,Digital Image Processing,2nd Edition.第44页/共56页Basic Relationship of Pixelsxy(0,0)Conventional indexing method(x,y)(x+1,y)(x-1,y)(x,y-1)(x,y+1)(x+1,y-1)(x-1,y-1)(x-1,y+1)(x+1,y+1)第45页/共56页Neighbors of a Pixelp(x+1,y)(x-1,y)(x,y-1)(x,y+1)4-neighbors o

30、f p:N4(p)=(x-1,y)(x+1,y)(x,y-1)(x,y+1)Neighborhood relation is used to tell adjacent pixels.It is useful for analyzing regions.Note:q N4(p)implies p N4(q)4-neighborhood relation considers only vertical and horizontal neighbors.第46页/共56页p(x+1,y)(x-1,y)(x,y-1)(x,y+1)(x+1,y-1)(x-1,y-1)(x-1,y+1)(x+1,y+1

31、)Neighbors of a Pixel(cont.)8-neighbors of p:(x-1,y-1)(x,y-1)(x+1,y-1)(x-1,y)(x+1,y)(x-1,y+1)(x,y+1)(x+1,y+1)N8(p)=8-neighborhood relation considers all neighbor pixels.第47页/共56页p(x+1,y-1)(x-1,y-1)(x-1,y+1)(x+1,y+1)Diagonal neighbors of p:ND(p)=(x-1,y-1)(x+1,y-1)(x-1,y+1)(x+1,y+1)Neighbors of a Pixe

32、l(cont.)Diagonal-neighborhood relation considers only diagonalneighbor pixels.第48页/共56页ConnectivityConnectivity is adapted from neighborhood relation.Two pixels are connected if they are in the same class(i.e.the same color or the same range of intensity)and they are neighbors of one another.For p a

33、nd q from the same classw 4-connectivity:p and q are 4-connected if q N4(p)w 8-connectivity:p and q are 8-connected if q N8(p)w mixed-connectivity(m-connectivity):p and q are m-connected if q N4(p)or q ND(p)and N4(p)N4(q)=第49页/共56页Adjacency A pixel p is adjacent to pixel q is they are connected.Two

34、image subsets S1 and S2 are adjacent if some pixelin S1 is adjacent to some pixel in S2S1S2We can define type of adjacency:4-adjacency,8-adjacencyor m-adjacency depending on type of connectivity.第50页/共56页Path A path from pixel p at(x,y)to pixel q at(s,t)is a sequenceof distinct pixels:(x0,y0),(x1,y1

35、),(x2,y2),(xn,yn)such that (x0,y0)=(x,y)and(xn,yn)=(s,t)and (xi,yi)is adjacent to(xi-1,yi-1),i=1,npq We can define type of path:4-path,8-path or m-pathdepending on type of adjacency.第51页/共56页Path(cont.)pqpqpq 8-path from p to qresults in some ambiguity m-path from p to qsolves this ambiguity8-pathm-

36、path第52页/共56页Distance For pixel p,q,and z with coordinates(x,y),(s,t)and(u,v),D is a distance function or metric ifw D(p,q)0(D(p,q)=0 if and only if p=q)w D(p,q)=D(q,p)w D(p,z)D(p,q)+D(q,z)Example:Euclidean distance第53页/共56页Distance(cont.)D4-distance(city-block distance)is defined as1210121222222Pixels with D4(p)=1 is 4-neighbors of p.第54页/共56页Distance(cont.)D8-distance(chessboard distance)is defined as1210121222222Pixels with D8(p)=1 is 8-neighbors of p.222222221111第55页/共56页感谢您的观看！第56页/共56页