DIP Notes Unit-1 PPT.pdf

DIP notes

1. 6CS3-01: Digital Image Processing
UNIT 1
Topic: Digital Image Representation, Sampling & Quantization,Steps in
Image Processing, Image acquisition, Color Image Representation
Presented By: Yashika Saini, Assistant Professor
Department of Computer Science Engineering
Arya Institute of Engineering & Technology, Jaipur

2. Introduction to Image Processing
•Imageisdefinedasa2-dimensionalfunction,
F(x,y),Wherex&yarespatialcoordinates,F
isintensityoramplitude.
•Animageisaprojectionofa3Dsceneontoa
2Dprojectionplane.
•TypesofImages:
1.AnalogImage
2.DigitalImage

3. •AnalogImage:Ananalogimagecanbe
representedasacontinousrangeofvalues
representingposition(x,y)andIntensity(F).
•DigitalImage:Whenx,y&Fareallfinite
discretequantities,imageisdigitalimage.
•Adigitalimageiscomposedofpicture
elementscalledpixels.
•DigitalImageProcessing:Theanalysisand
manipulationofadigitizedimage,inorderto
improveitsqualityiscalleddigitalimage
processing.

4. Applications of Digital Image Processing
1.Gamma-RayImaging:Majorusesofimagingbased
ongammaraysincludenuclearmedicineand
astronomicalobservations.

5. 2. X-Ray Imaging:
•X-raysareamongtheoldestsourcesofEMradiationusedfor
imaging.
•X-raysformedicalandindustrialimagingaregeneratedusing
anX-raytube,whichisavacuumtubewithacathodeand
anode

6. 3. Imaging in the Ultraviolet Band:
•Theyincludelithography,industrialinspection,
microscopy,lasers,biologicalimaging,and
astronomicalobservations.
•Ultravioletlightisusedinfluorescencemicroscopy,
oneofthefastestgrowingareasofmicroscopy.

7. 4. To improve Quality, remove noise etc.
5. Imaging in the Microwave Band:
•Theprincipalapplicationofimaginginthe
microwavebandisradar.
•Theuniquefeatureofimagingradarisitsabilityto
collectdataovervirtuallyanyregionatanytime,
regardlessofweatherorambientlightingconditions.

8. Components of Image Processing System

9. Image Representation
Before we discuss image acquisition recall that a digital
image is composed of Mrows and Ncolumns of
pixels each storing a value.
Pixel values are most
often grey levels in the
range 0-255(black-white).
We will see later on
that images can easily
be represented as
matrices
col
row
f (row, col)

10. Image Sensing
•Incomingenergylandsonasensormaterial
responsivetothattypeofenergyandthisgeneratesa
voltage.
•Collectionsofsensorsarearrangedtocaptureimages.
1. Single Sensor
2. Line Sensors
3. Array Sensors

11. Image Acquisition
Imagesaretypicallygeneratedbyilluminatinga
sceneandabsorbingtheenergyreflectedbythe
objectsinthatscene.
–Typical notions of
illumination and scene
can be way off:
•X-rays of a skeleton
•Ultrasound of an
unborn baby
•Electro-microscopic
images of molecules

12. Image Acquisition
Image Acquisition can be done in three ways:
1. Image Acquisition using Single Sensor.
2. Image Acquisition using Sensor Strip.
3. Image Acquisition using sensor arrays.

13. Images areanalog
•Noticethatwedefinedimagesasfunctionsina
continuousdomain.
•Imagesarerepresentationsofananalogworld.
•Hence,aswithalldigitalsignalprocessing,
weneedtodigitizeourimages.

14. Digitalization
•Digitalizationofananalogsignalinvolves
twooperations:
•Sampling,and
•Quantization
•Bothoperationscorrespondtodiscretizationof
aquantity,butindifferentdomains.

15. Image Sampling & Quantization
•Toconvertanalogimageintodigitalimagewehave
twointermediatesteps.
•Digitizingthecoordinatesvalues(x,y)iscalled
samplinganddigitizingtheamplitudevalues(F)is
calledquantization.

17. Representing Digital Images
•Letf(s,t)beacontinuousimagefunction,Wheres&t
arecontinuousvariables.
•Weconvertthisfunctionintoadigitalimageby
samplingandquantization.
•Supposethatwesamplethecontinuousimageintoa
2-Darray,f(x,y),havingMrowsandNcolumns,
where(x,y)arediscretecoordinates.
•Where, x = 0, 1, 2,….., M -1
y = 0, 1, 2,…..., N-1.

18. Representing Digital Images
•Therearethreebasicwaystorepresentf(x,y).

19. Digital Image Representation
•Figureshowsaplotofthefunction,withtwoaxesdetermining
spatiallocationandthethirdaxisbeingthevaluesoff
(intensities)asafunctionofthetwospatialvariablesxandy.
•Thistypeofrepresentationisnotpreferablebecauseindealing
withcompleximages,interpolationbecomesverydifficult.
Figure: Image plotted as a surface.

20. Digital Image Representation
•Itshowsf(x,y)asitwouldappearonamonitoror
photograph.Here,theintensityofeachpointis
proportionaltothevalueoffatthatpoint.
Figure: Image displayed as a visual intensity array

21. Digital Image Representation
•Inthisfigure,thereareonlythreediscrete
intensityvalues.Iftheintensityisnormalized
totheinterval[0,1],theneachpointinthe
imagehasthevalue0,0.5,or1.
•Amonitororprintersimplyconvertsthese
threevaluestoblack,gray,orwhite,
respectively.

22. Digital Image Representation
•Thethirdrepresentationissimplytodisplaythenumerical
valuesoff(x,y)asanarray(matrix).
•Whendevelopingalgorithms,thisrepresentationisquite
usefulwhenonlypartsoftheimageareprintedandanalyzed
asnumericalvalues.
Figure:2-Dnumericalarray(0,.5,
and1)representblack,gray,and
white.

23. Digital Image Representation
•Inequationform,wewritetherepresentation
ofanMxNnumericalarrayas
•Each element of this matrix is called an image
element, picture element, pixel or pel.

24. Digital Image Representation
•M&Nshouldbepositiveintegers.Butthe
numberofintensitylevelsLshould
•Where,Kisinteger.
•Bitsrequiredtostoreadigitizedimageis
b=MxNxK
WhenM=N(No.ofrows=No.ofcolumn),then

25. Steps in Image Processing

26. Image Acquisition
•Theimageiscapturedbyasensor(eg.Camera),and
digitizediftheoutputofthecameraorsensorisnot
alreadyindigitalform,usinganalogue-to-digital
convertor.

27. Image Enhancement
•Theprocessofmanipulatinganimagesothatthe
resultismoresuitablethantheoriginalfor
specificapplications.
•Theideabehindenhancementtechniquesisto
bringoutdetailsthatarehidden,orsimpleto
highlightcertainfeaturesofinterestinanimage.

28. Image Enhancement

29. Image Restoration
•Improvingtheappearanceofanimage.
•Tendtobemathematicalorprobabilistic
models.Enhancement,ontheotherhand,is
basedonhumansubjectivepreferences
regardingwhatconstitutesa“good”
enhancementresult.

30. Image Restoration

31. Color Image Processing
•Usethecoloroftheimagetoextractfeatures
ofinterestinanimage.
•Itincludescolormodelingandprocessingin
digitaldomain.

32. Wavelets & Multi-Resolution
•Waveletsaresmallwavesoflimitedduration
whichareusedtocalculatewavelettransform
whichprovidestimefrequencyinformation.
•Waveletsleadtomultiresolutionprocessingin
whichimagesarerepresentedinvarious
degreesofresolution.

33. Compression
•Techniquesforreducingthestoragerequiredto
saveanimageorthebandwidthrequiredto
transmitit.

34. Morphological Processing
•Toolsforextractingimagecomponentsthatare
usefulintherepresentationanddescriptionof
shape.
•Inthisstep,therewouldbeatransitionfrom
processesthatoutputimages,toprocessesthat
outputimageattributes.

35. Image Segmentation
•Segmentationprocedurespartitionanimage
intoitsconstituentpartsorobjects.
•Themoreaccuratethesegmentation,themore
likelyrecognitionistosucceed.
•Itisgenerallyusedtolocateobjectsand
boundariesinobjects.

36. Image Segmentation

37. Image Segmentation

38. Representation and Description
•Representation:Makeadecisionwhetherthe
datashouldberepresentedasaboundaryoras
acompleteregion.Itisalmostalwaysfollows
theoutputofasegmentationstage.
•BoundaryRepresentation:Focusonexternal
shapecharacteristics,suchascornersand
Inflections.
•RegionRepresentation:Focusoninternal
properties,suchastextureorskeletonShape.

39. Representation and Description
•Choosingarepresentationisonlypartofthesolutionfor
transformingrawdataintoaformsuitableforsubsequent
computerprocessing(mainlyrecognition).
•Description:alsocalled,featureselection,dealswith
extractingattributesthatresultinsomeinformationofinterest.

40. Object Recognition
•Recognition:Theprocessthatassignslabelto
anobjectbasedontheinformationprovidedby
itsdescription.

41. Knowledge Base
•Knowledgeaboutaproblemdomainiscoded
intoanimageprocessingsystemintheformof
aknowledgedatabase.

42. Spatial and Intensity Resolution
•Spatialresolutionisameasureofthesmallest
discernibledetailsinanimage.
•Itcanalsobestatedinnumberofways,dots(pixels)
perunitdistanceorlinepairsperunitdistance.
•Measuringspatialresolution
1.Dotsperinch
2.Linesperinch
3.Pixelsperinch

43. Measuring Spatial Resolution
1.Dots per inch
•DPIisameasureofimageresolution.
•DPImeansthathowmanydotsofinkareprintedperinch.
Whenanimagegetprintedoutfromprinter.
•ThehighertheDPIoftheprinterthehigheristhequalityof
theprintedimageorpaper.
•Commonlyusedinprintingandpublishingindustry.
•ForNewspapers:75dpi,Magazines:133dpi,Glossy
Brochure:175dpiandBookpages:2400dpi.

44. Measuring Spatial Resolution
2. Lines per inch
•LPIisusedtomeasuretheresolutionofimagesprintedin
halftones(Animagecomprisedofsuchdotsofonecoloris
usuallycalledahalftoneimage).Becausehalftoneimagesare
printedasaseriesofdots.
•LPIisameasurementofprintingresolution.
•LPIisusuallyusedinlaserprintersandgraphicdesign.

45. Measuring Spatial Resolution
3. Pixels per inch
•PPIrefersdisplayresolution,orhowmany
individualpixelsaredisplayedinoneinchofadigitalimage.
•PPIismeasurefordifferentdevicessuchastablets,mobile
phonesetc.
•ThehigheristhePPI,thehigheristhequality.

46. Intensity Resolution
•Intensityresolutionreferstothesmallest
discerniblechangeinintensitylevel.
•Intensityofresolutionmeansthenumberof
pixelspersquareinch,whichdeterminesthe
clarityorsharpnessofanimage.

47. Illustration of the effects of reducing image spatial resolution

48. Effects of varying the number of intensity levels in a digital image
Here Value of K varies from 8 to 1(8,7,6,5,4,3,2,1)

49. Image Interpolation
•Interpolationisatoolwhichisusedtoresizethe
imagesuchaszooming,shrinking,rotatingand
geometriccorrections.
•ImageInterpolationisalsocalledasre-samplingof
image.
•Inordertoresizetheimage,wehavetoresamplethe
image.
•ForExample:Supposethatanimageofsize500×
500pixelshastobeenlarged1.5timesto750×750
pixels.

50. Nearest Neighbor Interpolation
•Inthismethod,itassignstoeachnewlocationthe
intensityofitsnearestneighborintheoriginalimage.
•Bilinearinterpolation:Inwhichweusethefour
nearestneighborstoestimatetheintensityatagiven
location.
•Bicubicinterpolation:whichinvolvesthesixteen
nearestneighborsofapoint.

52. Introduction to Color Image Representation
•Colorisapowerfuldescriptorwhichsimplifiesobject
identificationandextractionfromascene.
•Humanbeingcanperceivefarmorehighernumberofcolor
shadesthangrayscaleshades.
•ColorImageProcessingisdividedintotwomajorcategories:
1.Fullcolor:Theimagesareacquiredwithafull-colorsensor,
suchasacolorTVorcolorscanner.
2.Pseudo-color:Assignacolortoaparticularrangeof
intensities.
Pseudocolorimagesaregrayscalewhichareassignedcolor
basedontheintensitiesvalues.

53. Full color & Pseudo-color Processing

56. Characterization of light
•IfthelightisAchromatic,itsonlyattributeis
itsintensity.Achromaticlightiswhatviewers
seeonablackandwhitetelevisionset.
•Chromaticlightspanstheelectromagnetic
spectrumfrom400to700nm.
•Threequantitiesdescribethequalityof
chromaticlight:radiance,luminanceand
brightness.

57. Characterization of light
•Radiance(Watts-W):Itisthetotalamountof
energycomingoutofthelightsource.
•Luminance(lumens-lm):Itgivesameasureof
amountofenergyanobserverperceivesfroma
lightsource.
•Brightness(nounit):Itissubjectivemeasure
thatispracticallyimpossibletomeasure.It
correspondstoachromaticattributeof
intensity.

58. Color Standardization
•CIE(CommissionInternationalEclairage)has
standardizedspecificwavelengthvaluestothree
primarycolors:
•Blue=435.8nm
•Green=546.1nm
•Red=700nm

59. Primary & Secondary Colors
•Theprimarycolorscanbeaddedtoproducethesecondary
colorsoflight-
oMagenta(redplusblue)
oCyan(greenplusblue)
oYellow(redplusgreen)
•Mixingthethreeprimariesorsecondarywithitsopposite
primarycolor,intherightintensitiesproduceswhitelight.

60. Primary & Secondary Colors
•Theprimarycolorsofpigmentsaremagenta,cyanand
yellowandthesecondarycolorsarered,greenandblue.
•Apropercombinationofthethreepigmentprimariesora
secondarywithitsoppositeprimaryproducesblack.

61. Characterizing Color
•Onecolorcanbedistinguishedfromotherbyusing
threecharacteristics.
•Brightness-Itembodiestheachromaticnotionof
intensity.
•Hue-Dominantcolor(wavelengthoflight)as
perceivedbyanobserver.Itismeasureofcolorofthe
object&itshueisexpressedasanangle.
•Saturation-Itreferstotherelativepurityorthe
amountofwhitelightmixedwithahue.OrItisa
measureoftherichnessofcolor.

62. Assignment Questions
Q1Definetheimageanddigitalimageprocessing.
Explainthefundamentalstepsofdigitalimage
processingwithsuitablediagram.
Q2Whataretheapplicationsofimageprocessing?
Explaincomponentsofimageprocessingsystem.
Q3Explainimagesensing&acquisition.
Q4ExplainimagesamplingandQuantizationprocessin
imageprocessing?
Q5Explaindigitalimagerepresentation.