DIP Quick Guide

10/26/2016
DIPQuickGuide
DIPQuickGuide
Advertisements
PreviousPage
NextPage
DigitalImageProcessingIntroduction
Introduction
Signalprocessingisadisciplineinelectricalengineeringandinmathematicsthatdealswith
analysisandprocessingofanaloganddigitalsignals,anddealswithstoring,filtering,and
other operations on signals. These signals include transmission signals , sound or voice
signals,imagesignals,andothersignalse.t.c.
Outofallthesesignals,thefieldthatdealswiththetypeofsignalsforwhichtheinputisan
imageandtheoutputisalsoanimageisdoneinimageprocessing.Asitnamesuggests,it
dealswiththeprocessingonimages.
Itcanbefurtherdividedintoanalogimageprocessinganddigitalimageprocessing.
Analogimageprocessing
Analog image processing is done on analog signals. It includes processing on two
dimensional analog signals. In this type of processing, the images are manipulated by
electrical means by varying the electrical signal. The common example include is the
televisionimage.
Digitalimageprocessinghasdominatedoveranalogimageprocessingwiththepassageof
timedueitswiderrangeofapplications.
Digitalimageprocessing
Thedigitalimageprocessingdealswithdevelopingadigitalsystemthatperformsoperations
onandigitalimage.
https://www.tutorialspoint.com/dip/dip_quick_guide.htm
1/179
10/26/2016
DIPQuickGuide
WhatisanImage
Animageisnothingmorethanatwodimensionalsignal.Itisdefinedbythemathematical
functionf(x,y)wherexandyarethetwocoordinateshorizontallyandvertically.
Thevalueoff(x,y)atanypointisgivesthepixelvalueatthatpointofanimage.
Theabovefigureisanexampleofdigitalimagethatyouarenowviewingonyourcomputer
screen.Butactually,thisimageisnothingbutatwodimensionalarrayofnumbersranging
between0and255.
128
30
123
232
123
321
123
77
89
80
255
255
Eachnumberrepresentsthevalueofthefunctionf(x,y)atanypoint.Inthiscasethevalue
128,230,123eachrepresentsanindividualpixelvalue.Thedimensionsofthepictureis
actuallythedimensionsofthistwodimensionalarray.
Relationshipbetweenadigitalimageandasignal
Iftheimageisatwodimensionalarraythenwhatdoesithavetodowithasignal?Inorder
tounderstandthat,Weneedtofirstunderstandwhatisasignal?
Signal
Inphysicalworld,anyquantitymeasurablethroughtimeoverspaceoranyhigherdimension
can be taken as a signal. A signal is a mathematical function, and it conveys some
information. A signal can be one dimensional or two dimensional or higher dimensional
signal.Onedimensionalsignalisasignalthatismeasuredovertime.Thecommonexample
isavoicesignal.Thetwodimensionalsignalsarethosethataremeasuredoversomeother
2/179
10/26/2016
DIPQuickGuide
physicalquantities.Theexampleoftwodimensionalsignalisadigitalimage.Wewilllookin
more detail in the next tutorial of how a one dimensional or two dimensional single and
highersignalsareformedandinterpreted.
Relationship
Sinceanythingthatconveysinformationorbroadcastamessageinphysicalworldbetween
twoobserversisasignal.Thatincludesspeechor(humanvoice)oranimageasasignal.
Sincewhenwespeak,ourvoiceisconvertedtoasoundwave/signalandtransformedwith
respect to the time to person we are speaking to. Not only this , but the way a digital
camera works, as while acquiring an image from a digital camera involves transfer of a
signalfromonepartofthesystemtotheother.
Howadigitalimageisformed
Since capturing an image from a camera is a physical process. The sunlight is used as a
source of energy. A sensor array is used for the acquisition of the image. So when the
sunlightfallsupontheobject,thentheamountoflightreflectedbythatobjectissensedby
thesensors,andacontinuousvoltagesignalisgeneratedbytheamountofsenseddata.In
order to create a digital image , we need to convert this data into a digital form. This
involves sampling and quantization. (They are discussed later on). The result of sampling
andquantizationresultsinantwodimensionalarrayormatrixofnumberswhicharenothing
butadigitalimage.
Overlappingfields
Machine/Computervision
Machinevisionorcomputervisiondealswithdevelopingasysteminwhichtheinputisan
image and the output is some information. For example: Developing a system that scans
humanfaceandopensanykindoflock.Thissystemwouldlooksomethinglikethis.
Computergraphics
3/179
10/26/2016
DIPQuickGuide
Computergraphicsdealswiththeformationofimagesfromobjectmodels,ratherthenthe
image is captured by some device. For example: Object rendering. Generating an image
fromanobjectmodel.Suchasystemwouldlooksomethinglikethis.
Artificialintelligence
Artificialintelligenceismoreorlessthestudyofputtinghumanintelligenceintomachines.
Artificial intelligence has many applications in image processing. For example: developing
computer aided diagnosis systems that help doctors in interpreting images of Xray , MRI
e.t.candthenhighlightingconspicuoussectiontobeexaminedbythedoctor.
Signalprocessing
Signal processing is an umbrella and image processing lies under it. The amount of light
reflected by an object in the physical world (3d world) is pass through the lens of the
cameraanditbecomesa2dsignalandhenceresultinimageformation.Thisimageisthen
digitized using methods of signal processing and then this digital image is manipulated in
digitalimageprocessing.
SignalsandSystemsIntroduction
This tutorial covers the basics of signals and system necessary for understanding the
conceptsofdigitalimageprocessing.Beforegoingintothedetailconcepts,letsfirstdefine
thesimpleterms.
Signals
In electrical engineering, the fundamental quantity of representing some information is
calledasignal.Itdoesnotmatterwhattheinformationisie:Analogordigitalinformation.
Inmathematics,asignalisafunctionthatconveyssomeinformation.Infactanyquantity
measurable through time over space or any higher dimension can be taken as a signal. A
signalcouldbeofanydimensionandcouldbeofanyform.
Analogsignals
4/179
10/26/2016
DIPQuickGuide
Asignalcouldbeananalogquantitythatmeansitisdefinedwithrespecttothetime.Itisa
continuous signal. These signals are defined over continuous independent variables. They
aredifficulttoanalyze,astheycarryahugenumberofvalues.Theyareverymuchaccurate
due to a large sample of values. In order to store these signals , you require an infinite
memorybecauseitcanachieveinfinitevaluesonarealline.Analogsignalsaredenotedby
sinwaves.
Forexample:
Humanvoice
Humanvoiceisanexampleofanalogsignals.Whenyouspeak,thevoicethatisproduced
travel through air in the form of pressure waves and thus belongs to a mathematical
function,havingindependentvariablesofspaceandtimeandavaluecorrespondingtoair
pressure.
Anotherexampleisofsinwavewhichisshowninthefigurebelow.
Y=sin(x)wherexisindepedent
Digitalsignals
As compared to analog signals, digital signals are very easy to analyze. They are
discontinuoussignals.Theyaretheappropriationofanalogsignals.
Theworddigitalstandsfordiscretevaluesandhenceitmeansthattheyusespecificvalues
to represent any information. In digital signal , only two values are used to represent
somethingie:1and0(binaryvalues).Digitalsignalsarelessaccuratethenanalogsignals
becausetheyarethediscretesamplesofananalogsignaltakenoversomeperiodoftime.
Howeverdigitalsignalsarenotsubjecttonoise.Sotheylastlongandareeasytointerpret.
Digitalsignalsaredenotedbysquarewaves.
5/179
10/26/2016
DIPQuickGuide
Forexample:
Computerkeyboard
Wheneverakeyispressedfromthekeyboard,theappropriateelectricalsignalissentto
keyboard controller containing the ASCII value that particular key. For example the
electricalsignalthatisgeneratedwhenkeyboardkeyaispressed,carryinformationofdigit
97intheformof0and1,whichistheASCIIvalueofcharactera.
Differencebetweenanaloganddigitalsignals
Comparison
element
Analogsignal
Digitalsignal
Analysis
Difficult
Possibletoanalyze
Representation
Continuous
Discontinuous
Accuracy
Moreaccurate
Lessaccurate
Storage
Infinitememory
Easilystored
SubjecttoNoise
Yes
No
Recording
Technique
Originalsignalispreserved
Samplesofthesignalaretakenand
preserved
Examples
Humanvoice,Thermometer,Analog
phonese.t.c
Computers,DigitalPhones,Digital
pens,e.t.c
Systems
Asystemisadefinedbythetypeofinputandoutputitdealswith.Sincewearedealingwith
signals , so in our case , our system would be a mathematical model , a piece of
code/software,oraphysicaldevice,orablackboxwhoseinputisasignalanditperforms
someprocessingonthatsignal,andtheoutputisasignal.Theinputisknownasexcitation
andtheoutputisknownasresponse.
6/179
10/26/2016
DIPQuickGuide
Intheabovefigureasystemhasbeenshownwhoseinputandoutputbotharesignalsbut
the input is an analog signal. And the output is an digital signal. It means our system is
actuallyaconversionsystemthatconvertsanalogsignalstodigitalsignals.
Letshavealookattheinsideofthisblackboxsystem
Conversionofanalogtodigitalsignals
Since there are lot of concepts related to this analog to digital conversion and viceversa.
Wewillonlydiscussthosewhicharerelatedtodigitalimageprocessing.Therearetwomain
conceptsthatareinvolvedinthecoversion.
Sampling
Quantization
Sampling
Samplingasitsnamesuggestscanbedefinedastakesamples.Takesamplesofadigital
signal over x axis. Sampling is done on an independent variable. In case of this
mathematicalequation:
Samplingisdoneonthexvariable.Wecanalsosaythattheconversionofxaxis(infinite
values)todigitalisdoneundersampling.
7/179
10/26/2016
DIPQuickGuide
Samplingisfurtherdivideintoupsamplinganddownsampling.Iftherangeofvaluesonx
axisarelessthenwewillincreasethesampleofvalues.Thisisknownasupsamplingand
itsviceversaisknownasdownsampling
Quantization
Quantization as its name suggest can be defined as dividing into quanta (partitions).
Quantizationisdoneondependentvariable.Itisoppositetosampling.
Incaseofthismathematicalequationy=sin(x)
Quantization is done on the Y variable. It is done on the y axis. The conversion of y axis
infinitevaluesto1,0,1(oranyotherlevel)isknownasQuantization.
These are the two basics steps that are involved while converting an analog signal to a
digitalsignal.
Thequantizationofasignalhasbeenshowninthefigurebelow.
Whydoweneedtoconvertananalogsignaltodigital
signal.
Thefirstandobviousreasonisthatdigitalimageprocessingdealswithdigitalimages,that
aredigitalsignals.Sowhenevertheimageiscaptured,itisconvertedintodigitalformat
andthenitisprocessed.
Thesecondandimportantreasonis,thatinordertoperformoperationsonananalogsignal
withadigitalcomputer,youhavetostorethatanalogsignalinthecomputer.Andinorder
8/179
10/26/2016
DIPQuickGuide
to store an analog signal , infinite memory is required to store it. And since thats not
possible,sothatswhyweconvertthatsignalintodigitalformatandthenstoreitindigital
computerandthenperformsoperationsonit.
Continuoussystemsvsdiscretesystems
Continuoussystems
Thetypeofsystemswhoseinputandoutputbotharecontinuoussignalsoranalogsignals
arecalledcontinuoussystems.
Discretesystems
Thetypeofsystemswhoseinputandoutputbotharediscretesignalsordigitalsignalsare
calleddigitalsystems
HistoryofPhotography
Originofcamera
9/179
10/26/2016
DIPQuickGuide
The history of camera and photography is not exactly the same. The concepts of camera
wereintroducedalotbeforetheconceptofphotography
CameraObscura
ThehistoryofthecameraliesinASIA.Theprinciplesofthecamerawerefirstintroducedby
aChinesephilosopherMOZI.Itisknownascameraobscura.Thecamerasevolvedfromthis
principle.
The word camera obscura is evolved from two different words. Camera and Obscura. The
meaningofthewordcameraisaroomorsomekindofvaultandObscurastandsfordark.
The concept which was introduced by the Chinese philosopher consist of a device, that
projectanimageofitssurroundingonthewall.HoweveritwasnotbuiltbytheChinese.
Thecreationofcameraobscura
TheconceptofChinesewasbringinrealitybyaMuslimscientistAbuAliAlHassanIbnal
HaithamcommonlyknownasIbnalHaitham.Hebuiltthefirstcameraobscura.Hiscamera
followstheprinciplesofpinholecamera.Hebuildthisdeviceinsomewherearound1000.
Portablecamera
In1685,afirstportablecamerawasbuiltbyJohannZahn.Beforetheadventofthisdevice,
thecameraconsistofasizeofroomandwerenotportable.Althoughadevicewasmadeby
anIrishscientistRobertBoyleandRobertHookethatwasatransportablecamera,butstill
thatdevicewasveryhugetocarryitfromoneplacetotheother.
Originofphotography
Althoughthecameraobscurawasbuiltin1000byaMuslimscientist.Butitsfirstactualuse
wasdescribedinthe13thcenturybyanEnglishphilosopherRogerBacon.Rogersuggested
10/179
10/26/2016
DIPQuickGuide
theuseofcamerafortheobservationofsolareclipses.
DaVinci
Although much improvement has been made before the 15th century , but the
improvementsandthefindingsdonebyLeonardodiserPierodaVinciwasremarkable.Da
Vinciwasagreatartist,musician,anatomist,andawarenginner.Heiscreditedformany
inventions.Hisoneofthemostfamouspaintingincludes,thepaintingofMonaLisa.
Davincinotonlybuiltacameraobscurafollowingtheprincipleofapinholecamerabutalso
usesitasdrawingaidforhisartwork.Inhiswork,whichwasdescribedinCodexAtlanticus
,manyprinciplesofcameraobscurahasbeendefined.
Hiscamerafollowstheprincipleofapinholecamerawhichcanbedescribedas
Whenimagesofilluminatedobjectspenetratethroughasmallholeintoaverydarkroom
youwillsee[ontheoppositewall]theseobjectsintheirproperformandcolor,reducedin
sizeinareversedposition,owingtotheintersectionofrays.
Firstphotograph
11/179
10/26/2016
DIPQuickGuide
Thefirstphotographwastakenin1814byaFrenchinventorJosephNicephoreNiepce.He
capturesthefirstphotographofaviewfromthewindowatLeGras,bycoatingthepewter
platewithbitumenandafterthatexposingthatplatetolight.
Firstunderwaterphotograph
ThefirstunderwaterphotographwastakenbyanEnglishmathematicianWilliamThomson
usingawatertightbox.Thiswasdonein1856.
Theoriginoffilm
The origin of film was introduced by an American inventor and a philanthropist known as
GeorgeEastmanwhoisconsideredasthepioneerofphotography.
HefoundedthecompanycalledasEastmanKodak,whichisfamousfordevelopingfilms.
Thecompanystartsmanufacturingpaperfilmin1885.HefirstcreatedthecameraKodak
andthenlaterBrownie.Browniewasaboxcameraandgainitspopularityduetoitsfeature
ofSnapshot.
12/179
10/26/2016
DIPQuickGuide
Aftertheadventofthefilm,thecameraindustryonceagaingotaboomandoneinvention
leadtoanother.
LeicaandArgus
Leicaandargusarethetwoanalogcamerasdevelopedin1925andin1939respectively.
ThecameraLeicawasbuiltusinga35mmcinefilm.
Argus was another camera analog camera that uses the 35mm format and was rather
inexpensiveascomparedbyLeicaandbecameverypopular.
13/179
10/26/2016
DIPQuickGuide
AnalogCCTVcameras
In1942aGermanengineerWalterBruchdevelopedandinstalledtheveryfirstsystemof
the analog CCTV cameras. He is also credited for the invention of color television in the
1960.
PhotoPac
Thefirstdisposablecamerawasintroducedin1949byPhotoPac.Thecamerawasonlya
onetimeusecamerawitharolloffilmalreadyincludedinit.ThelaterversionsofPhotopac
werewaterproofandevenhavetheflash.
DigitalCameras
MavicabySony
Mavica (the magnetic video camera) was launched by Sony in 1981 was the first game
changerindigitalcameraworld.Theimageswererecordedonfloppydisksandimagescan
beviewedlateronanymonitorscreen.
It was not a pure digital camera , but an analog camera. But got its popularity due to its
storingcapacityofimagesonafloppydisks.Itmeansthatyoucannowstoreimagesfora
long lasting period , and you can save a huge number of pictures on the floppy which are
replacedbythenewblankdisc,whentheygotfull.Mavicahasthecapacityofstoring25
imagesonadisk.
One more important thing that mavica introduced was its 0.3 mega pixel capacity of
capturingphotos.
14/179
10/26/2016
DIPQuickGuide
DigitalCameras
FujiDS1PcamerabyFujifilms1988wasthefirsttruedigitalcamera
Nikon D1 was a 2.74 mega pixel camera and the first commercial digital SLR camera
developedbyNikon,andwasverymuchaffordablebytheprofessionals.
Today digital cameras are included in the mobile phones with very high resolution and
quality.
ApplicationsandUsage
Since digital image processing has very wide applications and almost all of the technical
fieldsareimpactedbyDIP,wewilljustdiscusssomeofthemajorapplicationsofDIP.
DigitalImageprocessingisnotjustlimitedtoadjustthespatialresolutionoftheeveryday
images captured by the camera. It is not just limited to increase the brightness of the
photo,e.t.c.Ratheritisfarmorethanthat.
Electromagnetic waves can be thought of as stream of particles, where each particle is
moving with the speed of light. Each particle contains a bundle of energy. This bundle of
15/179
10/26/2016
DIPQuickGuide
energyiscalledaphoton.
Theelectromagneticspectrumaccordingtotheenergyofphotonisshownbelow.
In this electromagnetic spectrum, we are only able to see the visible spectrum. Visible
spectrum mainly includes seven different colors that are commonly term as (VIBGOYR).
VIBGOYRstandsforviolet,indigo,blue,green,orange,yellowandRed.
But that doesnot nullify the existence of other stuff in the spectrum. Our human eye can
onlyseethevisibleportion,inwhichwesawalltheobjects.Butacameracanseetheother
thingsthatanakedeyeisunabletosee.Forexample:xrays,gammarays,e.t.c.Hence
theanalysisofallthatstufftooisdoneindigitalimageprocessing.
Thisdiscussionleadstoanotherquestionwhichis
whydoweneedtoanalyzeallthatotherstuffinEM
spectrumtoo?
Theanswertothisquestionliesinthefact,becausethatotherstuffsuchasXRayhasbeen
widely used in the field of medical. The analysis of Gamma ray is necessary because it is
usedwidelyinnuclearmedicineandastronomicalobservation.Samegoeswiththerest of
thethingsinEMspectrum.
ApplicationsofDigitalImageProcessing
Some of the major fields in which digital image processing is widely used are mentioned
below
Imagesharpeningandrestoration
16/179
10/26/2016
DIPQuickGuide
Medicalfield
Remotesensing
Transmissionandencoding
Machine/Robotvision
Colorprocessing
Patternrecognition
Videoprocessing
MicroscopicImaging
Others
Imagesharpeningandrestoration
Image sharpening and restoration refers here to process images that have been captured
from the modern camera to make them a better image or to manipulate those images in
waytoachievedesiredresult.ItreferstodowhatPhotoshopusuallydoes.
This includes Zooming, blurring , sharpening , gray scale to color conversion, detecting
edgesandviceversa,ImageretrievalandImagerecognition.Thecommonexamplesare:
Theoriginalimage
Thezoomedimage
17/179
10/26/2016
DIPQuickGuide
Blurrimage
Sharpimage
Edges
18/179
10/26/2016
DIPQuickGuide
Medicalfield
ThecommonapplicationsofDIPinthefieldofmedicalis
1. Gammarayimaging
2. PETscan
3. XRayImaging
4. MedicalCT
5. UVimaging
UVimaging
Inthefieldofremotesensing,theareaoftheearthisscannedbyasatelliteorfromavery
highgroundandthenitisanalyzedtoobtaininformationaboutit.Oneparticularapplication
ofdigitalimageprocessinginthefieldofremotesensingistodetectinfrastructuredamages
causedbyanearthquake.
Asittakeslongertimetograspdamage,evenifseriousdamagesarefocusedon.Sincethe
areaeffectedbytheearthquakeissometimessowide,thatitnotpossibletoexamineit
withhumaneyeinordertoestimatedamages.Evenifitis,thenitisveryhecticandtime
consumingprocedure.Soasolutiontothisisfoundindigitalimageprocessing.Animageof
theeffectedareaiscapturedfromtheabovegroundandthenitisanalyzedtodetectthe
varioustypesofdamagedonebytheearthquake.
19/179
10/26/2016
DIPQuickGuide
Thekeystepsincludeintheanalysisare
1. Theextractionofedges
2. Analysisandenhancementofvarioustypesofedges
Transmissionandencoding
TheveryfirstimagethathasbeentransmittedoverthewirewasfromLondontoNewYork
viaasubmarinecable.Thepicturethatwassentisshownbelow.
Thepicturethatwassenttookthreehourstoreachfromoneplacetoanother.
Nowjustimagine,thattodayweareabletoseelivevideofeed,orlivecctvfootagefrom
onecontinenttoanotherwithjustadelayofseconds.Itmeansthatalotofworkhasbeen
doneinthisfieldtoo.Thisfielddoesnotonlyfocusontransmission,butalsoonencoding.
Manydifferentformatshavebeendevelopedforhighorlowbandwithtoencodephotosand
thenstreamitovertheinternetore.t.c.
Machine/Robotvision
Apartformthemanychallengesthatarobotfacetoday,oneofthebiggestchallengestillis
toincreasethevisionoftherobot.Makerobotabletoseethings,identifythem,identify
20/179
10/26/2016
DIPQuickGuide
thehurdlese.t.c.Muchworkhasbeencontributedbythisfieldandacompleteotherfieldof
computervisionhasbeenintroducedtoworkonit.
Hurdledetection
Hurdledetectionisoneofthecommontaskthathasbeendonethroughimageprocessing,
by identifying different type of objects in the image and then calculating the distance
betweenrobotandhurdles.
Linefollowerrobot
Mostoftherobotstodayworkbyfollowingthelineandthusarecalledlinefollowerrobots.
Thishelparobottomoveonitspathandperformsometasks.Thishasalsobeenachieved
throughimageprocessing.
Colorprocessing
Color processing includes processing of colored images and different color spaces that are
used.ForexampleRGBcolormodel,YCbCr,HSV.Italsoinvolvesstudyingtransmission,
storage,andencodingofthesecolorimages.
Patternrecognition
21/179
10/26/2016
DIPQuickGuide
Patternrecognitioninvolvesstudyfromimageprocessingandfromvariousotherfieldsthat
includesmachinelearning(abranchofartificialintelligence).Inpatternrecognition,image
processingisusedforidentifyingtheobjectsinanimagesandthenmachinelearningisused
totrainthesystemforthechangeinpattern.Patternrecognitionisusedincomputeraided
diagnosis,recognitionofhandwriting,recognitionofimagese.t.c
Videoprocessing
A video is nothing but just the very fast movement of pictures. The quality of the video
dependsonthenumberofframes/picturesperminuteandthequalityofeachframebeing
used. Video processing involves noise reduction , detail enhancement , motion detection ,
framerateconversion,aspectratioconversion,colorspaceconversione.t.c.
ConceptofDimensions
Wewilllookatthisexampleinordertounderstandtheconceptofdimension.
Consider you have a friend who lives on moon, and he wants to send you a gift on your
birthday present. He ask you about your residence on earth. The only problem is that the
courier service on moon doesnot understand the alphabetical address, rather it only
understandthenumericalcoordinates.Sohowdoyousendhimyourpositiononearth?
Thatswherecomestheconceptofdimensions.Dimensionsdefinetheminimumnumberof
pointsrequiredtopointapositionofanyparticularobjectwithinaspace.
Soletsgobacktoourexampleagaininwhichyouhavetosendyourpositiononearthto
your friend on moon. You send him three pair of coordinates. The first one is called
longitude,thesecondoneiscalledlatitude,andthethirdoneiscalledaltitude.
These three coordinates define your position on the earth. The first two defines your
location,andthethirdonedefinesyourheightabovethesealevel.
22/179
10/26/2016
DIPQuickGuide
So that means that only three coordinates are required to define your position on earth.
That means you live in world which is 3 dimensional. And thus this not only answers the
questionaboutdimension,butalsoanswersthereason,thatwhyweliveina3dworld.
Sincewearestudyingthisconceptinreferencetothedigitalimageprocessing,soweare
nowgoingtorelatethisconceptofdimensionwithanimage.
Dimensionsofimage
Soifweliveinthe3dworld,meansa3dimensionalworld,thenwhatarethedimensionsof
an image that we capture. An image is a two dimensional, thats why we also define an
image as a 2 dimensional signal. An image has only height and width. An image doesnot
havedepth.Justhavealookatthisimagebelow.
Ifyouwouldlookattheabovefigure,itshowsthatithasonlytwoaxiswhicharetheheight
andwidthaxis.Youcannotperceivedepthfromthisimage.Thatswhywesaythatanimage
istwodimensionalsignal.Butoureyeisabletoperceivethreedimensionalobjects,butthis
would be more explained in the next tutorial of how the camera works , and image is
perceived.
Thisdiscussionleadstosomeotherquestionsthathow3dimensionsystemsisformedfrom
2dimension.
Howdoestelevisionworks?
If we look the image above , we will see that it is a two dimensional image. In order to
convertitintothreedimension,weneedoneotherdimension.Letstaketimeasthethird
dimension,inthatcasewewillmovethistwodimensionalimageoverthethirddimension
time. The same concept that happens in television, that helps us perceive the depth of
differentobjectsonascreen.DoesthatmeanthatwhatcomesontheT.Vorwhatweseein
23/179
10/26/2016
DIPQuickGuide
thetelevisionscreenis3d.Wellwecanyes.Thereasonisthat,incaseofT.Vweifweare
playingavideo.Thenavideoisnothingelsebuttwodimensionalpicturesmoveovertime
dimension.Astwodimensionalobjectsaremovingoverthethirddimensionwhichisatime
sowecansayitis3dimensional.
Differentdimensionsofsignals
1dimensionsignal
The common example of a 1 dimension signal is a waveform. It can be mathematically
representedas
F(x)=waveform
Wherexisanindependentvariable.Sinceitisaonedimensionsignal,sothatswhythereis
onlyonevariablexisused.
Pictorialrepresentationofaonedimensionalsignalisgivenbelow:
Theabovefigureshowsaonedimensionalsignal.
Nowthisleadtoanotherquestion,whichis,eventhoughitisaonedimensionalsignal,then
why does it have two axis?. The answer to this question is that even though it is a one
dimensionalsignal,butwearedrawingitinatwodimensionalspace.Orwecansaythat
thespaceinwhichwearerepresentingthissignalistwodimensional.Thatswhyitlookslike
atwodimensionalsignal.
Perhaps you can understand the concept of one dimension more better by looking at the
figurebelow.
24/179
10/26/2016
DIPQuickGuide
Nowreferbacktoourinitialdiscussionondimension,Considertheabovefigurearealline
withpositivenumbersfromonepointtotheother.Nowifwehavetoexplainthelocationof
anypointonthisline,wejustneedonlyonenumber,whichmeansonlyonedimension.
2dimensionssignal
The common example of a two dimensional signal is an image , which has already been
discussedabove.
As we have already seen that an image is two dimensional signal, ie: it has two
dimensions.Itcanbemathematicallyrepresentedas:
F(x,y)=Image
Where x and y are two variables. The concept of two dimension can also be explained in
termsofmathematicsas:
Nowintheabovefigure,labelthefourcornersofthesquareasA,B,CandDrespectively.If
wecall,onelinesegmentinthefigureABandtheotherCD,thenwecanseethatthese
two parallel segments join up and make a square. Each line segment corresponds to one
dimension,sothesetwolinesegmentscorrespondto2dimensions.
3dimensionsignal
Three dimensional signal as it names refers to those signals which has three dimensions.
Themostcommonexamplehasbeendiscussedinthebeginningwhichisofourworld.We
25/179
10/26/2016
DIPQuickGuide
live in a three dimensional world. This example has been discussed very elaborately.
Anotherexampleofathreedimensionalsignalisacubeoravolumetricdataorthemost
commonexamplewouldbeanimatedor3dcartooncharacter.
Themathematicalrepresentationofthreedimensionalsignalis:
F(x,y,z)=animatedcharacter.
AnotheraxisordimensionZisinvolvedinathreedimension,thatgivestheillusionofdepth.
InaCartesiancoordinatesystemitcanbeviewedas:
4dimensionsignal
Inafourdimensionalsignal,fourdimensionsareinvolved.Thefirstthreearethesameas
ofthreedimensionalsignalwhichare:(X,Y,Z),andthefourthonewhichisaddedtothem
is T(time). Time is often referred to as temporal dimension which is a way to measure
change.Mathematicallyafourdsignalcanbestatedas:
F(x,y,z,t)=animatedmovie.
The common example of a 4 dimensional signal can be an animated 3d movie. As each
characterisa3dcharacterandthentheyaremovedwithrespecttothetime,duetowhich
wesawanillusionofathreedimensionalmoviemorelikearealworld.
Sothatmeansthatinrealitytheanimatedmoviesare4dimensionalie:movementof3d
charactersoverthefourthdimensiontime.
ImageFormationonCamera
Howhumaneyeworks?
26/179
10/26/2016
DIPQuickGuide
Before we discuss , the image formation on analog and digital cameras , we have to first
discusstheimageformationonhumaneye.Becausethebasicprinciplethatisfollowedby
thecamerashasbeentakenfromtheway,thehumaneyeworks.
When light falls upon the particular object , it is reflected back after striking through the
object.Theraysoflightwhenpassedthroughthelensofeye,formaparticularangle,and
the image is formed on the retina which is the back side of the wall. The image that is
formedisinverted.Thisimageistheninterpretedbythebrainandthatmakesusableto
understandthings.Duetoangleformation,weareabletoperceivetheheightanddepthof
the object we are seeing. This has been more explained in the tutorial of perspective
transformation.
Asyoucanseeintheabovefigure,thatwhensunlightfallsontheobject(inthiscasethe
objectisaface),itisreflectedbackanddifferentraysformdifferentanglewhentheyare
passed through the lens and an invert image of the object has been formed on the back
wall.Thelastportionofthefiguredenotesthattheobjecthasbeeninterpretedbythebrain
andreinverted.
Nowletstakeourdiscussionbacktotheimageformationonanaloganddigitalcameras.
Imageformationonanalogcameras
27/179
10/26/2016
DIPQuickGuide
Inanalogcameras,theimageformationisduetothechemicalreactionthattakesplaceon
thestripthatisusedforimageformation.
A35mmstripisusedinanalogcamera.Itisdenotedinthefigureby35mmfilmcartridge.
Thisstripiscoatedwithsilverhalide(achemicalsubstance).
A35mmstripisusedinanalogcamera.Itisdenotedinthefigureby35mmfilmcartridge.
Thisstripiscoatedwithsilverhalide(achemicalsubstance).
Lightisnothingbutjustthesmallparticlesknownasphotonparticles.Sowhenthesephoton
particlesarepassedthroughthecamera,itreactswiththesilverhalideparticlesonthestrip
anditresultsinthesilverwhichisthenegativeoftheimage.
Inordertounderstanditbetter,havealookatthisequation.
Photons(lightparticles)+silverhalide?silver?imagenegative.
28/179
10/26/2016
DIPQuickGuide
This is just the basics, although image formation involves many other concepts regarding
thepassingoflightinside,andtheconceptsofshutterandshutterspeedandapertureand
itsopeningbutfornowwewillmoveontothenextpart.Althoughmostoftheseconcepts
havebeendiscussedinourtutorialofshutterandaperture.
This is just the basics, although image formation involves many other concepts regarding
thepassingoflightinside,andtheconceptsofshutterandshutterspeedandapertureand
itsopeningbutfornowwewillmoveontothenextpart.Althoughmostoftheseconcepts
havebeendiscussedinourtutorialofshutterandaperture.
Imageformationondigitalcameras
Inthedigitalcameras,theimageformationisnotduetothechemicalreactionthattake
place , rather it is a bit more complex then this. In the digital camera , a CCD array of
sensorsisusedfortheimageformation.
ImageformationthroughCCDarray
CCD stands for chargecoupled device. It is an image sensor, and like other sensors it
senses the values and converts them into an electric signal. In case of CCD it senses the
imageandconvertitintoelectricsignale.t.c.
29/179
10/26/2016
DIPQuickGuide
ThisCCDisactuallyintheshapeofarrayorarectangulargrid.Itislikeamatrixwitheach
cellinthematrixcontainsacensorthatsensestheintensityofphoton.
Like analog cameras , in the case of digital too , when light falls on the object , the light
reflectsbackafterstrikingtheobjectandallowedtoenterinsidethecamera.
EachsensoroftheCCDarrayitselfisananalogsensor.Whenphotonsoflightstrikeonthe
chip , it is held as a small electrical charge in each photo sensor. The response of each
sensorisdirectlyequaltotheamountoflightor(photon)energystrikedonthesurfaceof
thesensor.
Since we have already define an image as a two dimensional signal and due to the two
dimensionalformationoftheCCDarray,acompleteimagecanbeachievedfromthisCCD
array.
Ithaslimitednumberofsensors,anditmeansalimiteddetailcanbecapturedbyit.Also
eachsensorcanhaveonlyonevalueagainsttheeachphotonparticlethatstrikeonit.
So the number of photons striking(current) are counted and stored. In order to measure
accuratelythese,externalCMOSsensorsarealsoattachedwithCCDarray.
Introductiontopixel
ThevalueofeachsensoroftheCCDarrayreferstoeachthevalueoftheindividualpixel.
Thenumberofsensors=numberofpixels.Italsomeansthateachsensorcouldhaveonly
oneandonlyonevalue.
Storingimage
ThechargesstoredbytheCCDarrayareconvertedtovoltageonepixelatatime.Withthe
helpofadditionalcircuits,thisvoltageisconvertedintoadigitalinformationandthenitis
stored.
Eachcompanythatmanufacturesdigitalcamera,maketheirownCCDsensors.Thatinclude
,Sony,Mistubishi,Nikon,Samsung,Toshiba,FujiFilm,Canone.t.c.
30/179
10/26/2016
DIPQuickGuide
Apartfromtheotherfactors,thequalityoftheimagecapturedalsodependsonthetype
andqualityoftheCCDarraythathasbeenused.
CameraMechansim
Inthistutorial,wewilldiscusssomeofthebasiccameraconcepts,likeaperture,shutter,
shutter speed , ISO and we will discuss the collective use of these concepts to capture a
goodimage.
Aperture
Apertureisasmallopeningwhichallowsthelighttotravelinsideintocamera.Hereisthe
pictureofaperture.
You will see some small blades like stuff inside the aperture. These blades create a
octagonalshapethatcanbeopenedclosed.Andthusitmakesensethat,themoreblades
willopen,theholefromwhichthelightwouldhavetopasswouldbebigger.Thebiggerthe
hole,themorelightisallowedtoenter.
Effect
The effect of the aperture directly corresponds to brightness and darkness of an image. If
theapertureopeningiswide,itwouldallowmorelighttopassintothecamera.Morelight
wouldresultinmorephotons,whichultimatelyresultinabrighterimage.
Theexampleofthisisshownbelow
Considerthesetwophotos
31/179
10/26/2016
DIPQuickGuide
Theoneontherightsidelooksbrighter,itmeansthatwhenitwascapturedbythecamera,
theaperturewaswideopen.Ascomparetotheotherpictureontheleftside,whichisvery
dark as compare to the first one, that shows that when that image was captured, its
aperturewasnotwideopen.
Size
Nowletsdiscussthemathsbehindtheaperture.Thesizeoftheapertureisdenotedbyaf
value.Anditisinverselyproportionaltotheopeningofaperture.
Herearethetwoequations,thatbestexplainthisconcept.
Largeaperturesize=Smallfvalue
Smallaperturesize=Greaterfvalue
Pictoriallyitcanberepresentedas:
Shutter
32/179
10/26/2016
DIPQuickGuide
After the aperture , there comes the shutter. The light when allowed to pass from the
aperture,fallsdirectlyontotheshutter.Shutterisactuallyacover,aclosedwindow,or
can be thought of as a curtain. Remember when we talk about the CCD array sensor on
whichtheimageisformed.Wellbehindtheshutteristhesensor.Soshutteristheonlything
thatisbetweentheimageformationandthelight,whenitispassedfromaperture.
Assoonastheshutterisopen,lightfallsontheimagesensor,andtheimageisformedon
thearray.
Effect
If the shutter allows light to pass a bit longer , the image would be brighter. Similarly a
darkerpictureisproduced,whenashutterisallowedtomoveveryquicklyandhence,the
light that is allowed to pass has very less photons , and the image that is formed on the
CCDarraysensorisverydark.
Shutterhasfurthertwomainconcepts:
ShutterSpeed
Shuttertime
Shutterspeed
Theshutterspeedcanbereferredtoasthenumberoftimestheshuttergetopenorclose.
Rememberwearenottalkingaboutforhowlongtheshuttergetopenorclose.
Shuttertime
Theshuttertimecanbedefinedas
When the shutter is open , then the amount of wait time it take till it is closed is called
shuttertime.
Inthiscasewearenottalkingabouthowmanytimes,theshuttergotopenorclose,but
wearetalkingaboutforhowmuchtimedoesitremainwideopen.
Forexample:
Wecanbetterunderstandthesetwoconceptsinthisway.Thatletssaythatashutteropens
15timesandthengetclosed,andforeachtimeitopensfor1secondandthengetclosed.
Inthisexample,15istheshutterspeedand1secondistheshuttertime.
Relationship
The relationship between shutter speed and shutter time is that they are both inversely
proportionaltoeachother.
33/179
10/26/2016
DIPQuickGuide
Thisrelationshipcanbedefinedintheequationbelow.
Moreshutterspeed=lessshuttertime
Lessshutterspeed=moreshuttertime.
Explanation:
Thelesserthetimerequired,themoreisthespeed.Andthegreaterthetimerequired,the
lessisthespeed.
Applications
Thesetwoconceptstogethermakeavarietyofapplications.Someofthemaregivenbelow.
Fastmovingobjects:
Ifyouweretocapturetheimageofafastmovingobject,couldbeacaroranything.The
adjustmentofshutterspeedanditstimewouldeffectalot.
So,inordertocaptureanimagelikethis,wewillmaketwoamendments:
Increaseshutterspeed
Decreaseshuttertime
Whathappensis,thatwhenweincreaseshutterspeed,themorenumberoftimes,the
shutterwouldopenorclose.Itmeansdifferentsamplesoflightwouldallowtopassin.And
when we decrease shutter time , it means we will immediately captures the scene, and
closetheshuttergate.
Ifyouwilldothis,yougetacrispimageofafastmovingobject.
In order to understand it , we will look at this example. Suppose you want to capture the
imageoffastmovingwaterfall.
Yousetyourshutterspeedto1secondandyoucaptureaphoto.Thisiswhatyouget
34/179
10/26/2016
DIPQuickGuide
Thenyousetyourshutterspeedtoafasterspeedandyouget.
Thenagainyousetyourshutterspeedtoevenmorefasterandyouget.
Youcanseeinthelastpicture,thatwehaveincreaseourshutterspeedtoveryfast,that
means that a shutter get opened or closed in 200th of 1 second and so we got a crisp
image.
ISO
ISO factor is measured in numbers. It denotes the sensitivity of light to camera. If ISO
numberislowered,itmeansourcameraislesssensitivetolightandiftheISOnumberis
high,itmeansitismoresenstivie.
Effect
ThehigheristheISO,themorebrighterthepicturewouldbe.IFISOissetto1600,the
picturewouldbeverybrighterandviceversa.
Sideeffect
If the ISO increases, the noise in the image also increases. Today most of the camera
manufacturingcompaniesareworkingonremovingthenoisefromtheimagewhenISOis
settohigherspeed.
ConceptofPixel
35/179
10/26/2016
DIPQuickGuide
ConceptofPixel
Pixel
Pixelisthesmallestelementofanimage.Eachpixelcorrespondtoanyonevalue.Inan8
bitgrayscaleimage,thevalueofthepixelbetween0and255.Thevalueofapixelatany
pointcorrespondtotheintensityofthelightphotonsstrikingatthatpoint.Eachpixelstorea
valueproportionaltothelightintensityatthatparticularlocation.
PEL
A pixel is also known as PEL. You can have more understanding of the pixel from the
picturesgivenbelow.
Intheabovepicture,theremaybethousandsofpixels,thattogethermakeupthisimage.
We will zoom that image to the extent that we are able to see some pixels division. It is
shownintheimagebelow.
Intheabovepicture,theremaybethousandsofpixels,thattogethermakeupthisimage.
We will zoom that image to the extent that we are able to see some pixels division. It is
shownintheimagebelow.
RelationshipwithCCDarray
36/179
10/26/2016
DIPQuickGuide
WehaveseenthathowanimageisformedintheCCDarray.Soapixelcanalsobedefined
as
ThesmallestdivisiontheCCDarrayisalsoknownaspixel.
EachdivisionofCCDarraycontainsthevalueagainsttheintensityofthephotonstrikingto
it.Thisvaluecanalsobecalledasapixel
Calculationoftotalnumberofpixels
We have define an image as a two dimensional signal or matrix. Then in that case the
numberofPELwouldbeequaltothenumberofrowsmultiplywithnumberofcolumns.
Thiscanbemathematicallyrepresentedasbelow:
Totalnumberofpixels=numberofrows(X)numberofcolumns
Orwecansaythatthenumberof(x,y)coordinatepairsmakeupthetotalnumberofpixels.
We will look in more detail in the tutorial of image types , that how do we calculate the
pixelsinacolorimage.
Graylevel
Thevalueofthepixelatanypointdenotestheintensityofimageatthatlocation,andthat
isalsoknownasgraylevel.
Wewillseeinmoredetailaboutthevalueofthepixelsintheimagestorageandbitsper
pixeltutorial,butfornowwewilljustlookattheconceptofonlyonepixelvalue.
Pixelvalue.(0)
Asithasalreadybeendefineinthebeginningofthistutorial,thateachpixelcanhaveonly
onevalueandeachvaluedenotestheintensityoflightatthatpointoftheimage.
Wewillnowlookataveryuniquevalue0.Thevalue0meansabsenceoflight.Itmeans
that0denotesdark,anditfurthermeansthatwheneverapixelhasavalueof0,itmeans
atthatpoint,blackcolorwouldbeformed.
37/179
10/26/2016
DIPQuickGuide
Havealookatthisimagematrix
0
Nowthisimagematrixhasallfilledupwith0.Allthepixelshaveavalueof0.Ifwewereto
calculatethetotalnumberofpixelsformthismatrix,thisishowwearegoingtodoit.
Totalnoofpixels=totalno.ofrowsXtotalno.ofcolumns
=3X3
=9.
It means that an image would be formed with 9 pixels, and that image would have a
dimensionof3rowsand3columnandmostimportantlythatimagewouldbeblack.
Theresultingimagethatwouldbemadewouldbesomethinglikethis
Nowwhyisthisimageallblack.Becauseallthepixelsintheimagehadavalueof0.
PerspectiveTransformation
Whenhumaneyesseenearthingstheylookbiggerascomparetothosewhoarefaraway.
This is called perspective in a general way. Whereas transformation is the transfer of an
objecte.t.cfromonestatetoanother.
So overall , the perspective transformation deals with the conversion of 3d world into 2d
image. The same principle on which human vision works and the same principle on which
thecameraworks.
Wewillseeindetailaboutwhythishappens,thatthoseobjectswhichareneartoyoulook
bigger,whilethosewhoarefaraway,looksmallereventhoughtheylookbiggerwhenyou
reachthem.
Wewillstartthisdiscussionbytheconceptofframeofreference:
Frameofreference:
Frameofreferenceisbasicallyasetofvaluesinrelationtowhichwemeasuresomething.
38/179
10/26/2016
DIPQuickGuide
5framesofreference
Inordertoanalyzea3dworld/image/scene,5differentframeofreferencesarerequired.
Object
World
Camera
Image
Pixel
Objectcoordinateframe
Objectcoordinateframeisusedformodelingobjects.Forexample,checkingifaparticular
objectisinaproperplacewithrespecttotheotherobject.Itisa3dcoordinatesystem.
Worldcoordinateframe
Worldcoordinateframeisusedforcorelatingobjectsina3dimensionalworld.Itisa3d
coordinatesystem.
Cameracoordinateframe
Cameracoordinateframeisusedtorelateobjectswithrespectofthecamera.Itisa3d
coordinatesystem.
Imagecoordinateframe
It is not a 3d coordinate system , rather it is a 2d system. It is used to describe how 3d
pointsaremappedina2dimageplane.
Pixelcoordinateframe
Itisalsoa2dcoordinatesystem.Eachpixelhasavalueofpixelcoordinates.
39/179
10/26/2016
DIPQuickGuide
Transformationbetweenthese5frames
Thatshowa3dsceneistransformedinto2d,withimageofpixels.
Nowwewillexplainthisconceptmathematically.
Where
Y=3dobject
y=2dImage
f=focallengthofthecamera
Z=distancebetweenimageandthecamera
NowtherearetwodifferentanglesformedinthistransformwhicharerepresentedbyQ.
40/179
10/26/2016
DIPQuickGuide
Thefirstangleis
Whereminusdenotesthatimageisinverted.Thesecondanglethatisformedis:
Comparingthesetwoequationsweget
Fromthisequation,wecanseethatwhentheraysoflightreflectbackafterstrikingfrom
theobject,passedfromthecamera,aninvertimageisformed.
Wecanbetterunderstandthis,withthisexample.
Forexample
Calculatingthesizeofimageformed
Supposeanimagehasbeentakenofaperson5mtall,andstandingatadistanceof50m
fromthecamera,andwehavetotellthatwhatisthesizeoftheimageoftheperson,with
acameraoffocallengthis50mm.
Solution:
Sincethefocallengthisinmillimeter,sowehavetoconverteverythinginmillimeterin
ordertocalculateit.
So,
Y=5000mm.
f=50mm.
Z=50000mm.
Puttingthevaluesintheformula,weget
41/179
10/26/2016
DIPQuickGuide
=5mm.
Again,theminussignindicatesthattheimageisinverted.
ConceptofBitsPerPixel
Bpporbitsperpixeldenotesthenumberofbitsperpixel.Thenumberofdifferentcolorsin
animageisdependsonthedepthofcolororbitsperpixel.
Bitsinmathematics:
Itsjustlikeplayingwithbinarybits.
Howmanynumberscanberepresentedbyonebit.
0
1
Howmanytwobitscombinationscanbemade.
00
01
10
11
Ifwedeviseaformulaforthecalculationoftotalnumberofcombinationsthatcanbemade
frombit,itwouldbelikethis.
Wherebppdenotesbitsperpixel.Put1intheformulayouget2,put2intheformula,you
get4.Itgrowsexponentionally.
Numberofdifferentcolors:
Now as we said it in the beginning , that the number of different colors depend on the
numberofbitsperpixel.
42/179
10/26/2016
DIPQuickGuide
Thetableforsomeofthebitsandtheircolorisgivenbelow.
Bitsperpixel
Numberofcolors
1bpp
2colors
2bpp
4colors
3bpp
8colors
4bpp
16colors
5bpp
32colors
6bpp
64colors
7bpp
128colors
8bpp
256colors
10bpp
1024colors
16bpp
65536colors
24bpp
16777216colors(16.7millioncolors)
32bpp
4294967296colors(4294millioncolors)
Thistableshowsdifferentbitsperpixelandtheamountofcolortheycontain.
Shades
Youcaneasilynoticethepatternoftheexponentionalgrowth.Thefamousgrayscaleimage
isof8bpp,meansithas256differentcolorsinitor256shades.
Shadescanberepresentedas:
Colorimagesareusuallyofthe24bppformat,or16bpp.
Wewillseemoreaboutothercolorformatsandimagetypesinthetutorialofimagetypes.
Colorvalues:
Wehavepreviouslyseeninthetutorialofconceptofpixel,that0pixelvaluedenotesblackcolor.
Blackcolor:
43/179
10/26/2016
DIPQuickGuide
Remember,0pixelvaluealwaysdenotesblackcolor.Butthereisnofixedvaluethatdenotes
whitecolor.
Whitecolor:
Thevaluethatdenoteswhitecolorcanbecalculatedas:
Incaseof1bpp,0denotesblack,and1denoteswhite.
Incase8bpp,0denotesblack,and255denoteswhite.
Graycolor:
Whenyoucalculatetheblackandwhitecolorvalue,thenyoucancalculatethepixelvalue
ofgraycolor.
Graycolorisactuallythemidpointofblackandwhite.Thatsaid,
Incaseof8bpp,thepixelvaluethatdenotesgraycoloris127or128bpp(ifyoucountfrom
1,notfrom0).
Imagestoragerequirements
Afterthediscussionofbitsperpixel,nowwehaveeverythingthatweneedtocalculatea
sizeofanimage.
Imagesize
Thesizeofanimagedependsuponthreethings.
Numberofrows
Numberofcolumns
Numberofbitsperpixel
Theformulaforcalculatingthesizeisgivenbelow.
Sizeofanimage=rows*cols*bpp
Itmeansthatifyouhaveanimage,letssaythisone:
44/179
10/26/2016
DIPQuickGuide
Assumingithas1024rowsandithas1024columns.Andsinceitisagrayscaleimage,it
has 256 different shades of gray or it has bits per pixel. Then putting these values in the
formula,weget
Sizeofanimage=rows*cols*bpp
=1024*1024*8
=8388608bits.
Butsinceitsnotastandardanswerthatwerecognize,sowillconvertitintoourformat.
Convertingitintobytes=8388608/8=1048576bytes.
Convertingintokilobytes=1048576/1024=1024kb.
ConvertingintoMegabytes=1024/1024=1Mb.
Thatshowanimagesizeiscalculatedanditisstored.Nowintheformula,ifyouaregiven
thesizeofimageandthebitsperpixel,youcanalsocalculatetherowsandcolumnsofthe
image,providedtheimageissquare(samerowsandsamecolumn).
TypesofImages
Therearemanytypeofimages,andwewilllookindetailaboutdifferenttypesofimages,
andthecolordistributioninthem.
Thebinaryimage
Thebinaryimageasitnamestates,containonlytwopixelvalues.
0and1.
In our previous tutorial of bits per pixel , we have explained this in detail about the
representationofpixelvaluestotheirrespectivecolors.
Here0referstoblackcolorand1referstowhitecolor.ItisalsoknownasMonochrome.
45/179
10/26/2016
DIPQuickGuide
Blackandwhiteimage:
Theresultingimagethatisformedhenceconsistofonlyblackandwhitecolorandthuscan
alsobecalledasBlackandWhiteimage.
Nograylevel
Oneoftheinterestingthisaboutthisbinaryimagethatthereisnograylevelinit.Onlytwo
colorsthatareblackandwhitearefoundinit.
Format
BinaryimageshaveaformatofPBM(Portablebitmap)
2,3,4,5,6bitcolorformat
Theimageswithacolorformatof2,3,4,5and6bitarenotwidelyusedtoday.Theywere
usedinoldtimesforoldTVdisplays,ormonitordisplays.
Buteachofthesecolorshavemorethentwograylevels,andhencehasgraycolorunlike
thebinaryimage.
In a 2 bit 4, in a 3 bit 8 , in a 4 bit 16, in a 5 bit 32, in a 6 bit 64 different colors are
present.
8bitcolorformat
8bitcolorformatisoneofthemostfamousimageformat.Ithas256differentshadesof
colorsinit.ItiscommonlyknownasGrayscaleimage.
Therangeofthecolorsin8bitvaryfrom0255.Where0standsforblack,and255stands
forwhite,and127standsforgraycolor.
ThisformatwasusedinitiallybyearlymodelsoftheoperatingsystemsUNIXandtheearly
colorMacintoshes.
AgrayscaleimageofEinsteinisshownbelow:
46/179
10/26/2016
DIPQuickGuide
Format
TheformatoftheseimagesarePGM(PortableGrayMap).
Thisformatisnotsupportedbydefaultfromwindows.Inordertoseegrayscaleimage,you
needtohaveanimageviewerorimageprocessingtoolboxsuchasMatlab.
Behindgrayscaleimage:
Aswehaveexplaineditseveraltimesintheprevioustutorials,thatanimageisnothingbut
atwodimensionalfunction,andcanberepresentedbyatwodimensionalarrayormatrix.
So in the case of the image of Einstein shown above , there would be two dimensional
matrixinbehindwithvaluesrangingbetween0and255.
Butthatsnotthecasewiththecolorimages.
16bitcolorformat
Itisacolorimageformat.Ithas65,536differentcolorsinit.ItisalsoknownasHighcolor
format.
IthasbeenusedbyMicrosoftintheirsystemsthatsupportmorethen8bitcolorformat.
Now in this 16 bit format and the next format we are going to discuss which is a 24 bit
formatarebothcolorformat.
Thedistributionofcolorinacolorimageisnotassimpleasitwasingrayscaleimage.
A16bitformatisactuallydividedintothreefurtherformatswhichareRed,GreenandBlue.
Thefamous(RGB)format.
Itispictoriallyrepresentedintheimagebelow.
47/179
10/26/2016
DIPQuickGuide
Nowthequestionarises,thathowwouldyoudistribute16intothree.Ifyoudoitlikethis,
5bitsforR,5bitsforG,5bitsforB
Thenthereisonebitremainsintheend.
Sothedistributionof16bithasbeendonelikethis.
5bitsforR,6bitsforG,5bitsforB.
Theadditionalbitthatwasleftbehindisaddedintothegreenbit.Becausegreenisthecolor
whichismostsoothingtoeyesinallofthesethreecolors.
Notethisisdistributionisnotfollowedbyallthesystems.Somehaveintroducedanalpha
channelinthe16bit.
Anotherdistributionof16bitformatislikethis:
4bitsforR,4bitsforG,4bitsforB,4bitsforalphachannel.
Orsomedistributeitlikethis
5bitsforR,5bitsforG,5bitsforB,1bitsforalphachannel.
24bitcolorformat
24 bit color format also known as true color format. Like 16 bit color format , in a 24 bit
colorformat,the24bitsareagaindistributedinthreedifferentformatsofRed,Greenand
Blue.
48/179
10/26/2016
DIPQuickGuide
Since24isequallydividedon8,soithasbeendistributedequallybetweenthreedifferent
colorchannels.
Theirdistributionislikethis.
8bitsforR,8bitsforG,8bitsforB.
Behinda24bitimage.
Unlikea8bitgrayscaleimage,whichhasonematrixbehindit,a24bitimagehasthree
differentmatricesofR,G,B.
Format
Itisthemostcommonusedformat.ItsformatisPPM(PortablepixMap)whichissupported
by Linux operating system. The famous windows has its own format for it which is BMP (
Bitmap).
ColorCodesConversion
Inthistutorial,wewillseethathowdifferentcolorcodescanbecombinedtomakeother
colors,andhowwecancovertRGBcolorcodestohexandviceversa.
Differentcolorcodes
Allthecolorshereareofthe24bitformat,thatmeanseachcolorhas8bitsofred,8bits
ofgreen,8bitsofblue,init.Orwecansayeachcolorhasthreedifferentportions.Youjust
49/179
10/26/2016
DIPQuickGuide
havetochangethequantityofthesethreeportionstomakeanycolor.
Binarycolorformat
Color:Black
Image:
DecimalCode:
(0,0,0)
Explanation:
Asithasbeenexplainedintheprevioustutorials,thatinan8bitformat,0referstoblack.
Soifwehavetomakeapureblackcolor,wehavetomakeallthethreeportionofR,G,B
to0.
Color:White
Image:
DecimalCode:
(255,255,255)
Explanation:
SinceeachportionofR,G,Bisan8bitportion.Soin8bit,thewhitecolorisformedby255.
Itisexplainedinthetutorialofpixel.Soinordertomakeawhitecolorweseteachportion
to 255 and thats how we got a white color. By setting each of the value to 255 , we get
overallvalueof255,thatsmakethecolorwhite.
RGBcolormodel:
Color:Red
Image:
DecimalCode:
(255,0,0)
Explanation:
50/179
10/26/2016
DIPQuickGuide
Sinceweneedonlyredcolor,sowezeroouttherestofthetwoportionswhicharegreen
andblue,andwesettheredportiontoitsmaximumwhichis255.
Color:Green
Image:
DecimalCode:
(0,255,0)
Explanation:
Sinceweneedonlygreencolor,sowezeroouttherestofthetwoportionswhicharered
andblue,andwesetthegreenportiontoitsmaximumwhichis255.
Color:Blue
Image:
DecimalCode:
(0,0,255)
Explanation:
Sinceweneedonlybluecolor,sowezeroouttherestofthetwoportionswhichareredand
green,andwesettheblueportiontoitsmaximumwhichis255
Graycolor:
Color:Gray
Image:
DecimalCode:
(128,128,128)
Explanation:
Aswehavealreadydefinedinourtutorialofpixel,thatgraycolorIsactuallythemidpoint.
In an 8bit format , the mid point is 128 or 127. In this case we choose 128. So we set
eachoftheportiontoitsmidpointwhichis128,andthatresultsinoverallmidvalueand
wegotgraycolor.
51/179
10/26/2016
DIPQuickGuide
CMYKcolormodel:
CMYKisanothercolormodelwherecstandsforcyan,mstandsformagenta,ystandsfor
yellow,andkforblack.CMYKmodeliscommonlyusedincolorprintersinwhichthereare
twocartersofcolorisused.OneconsistofCMYandotherconsistofblackcolor.
ThecolorsofCMYcanalsomadefromchangingthequantityorportionofred,greenand
blue.
Color:Cyan
Image:
DecimalCode:
(0,255,255)
Explanation:
CyancolorisformedfromthecombinationoftwodifferentcolorswhichareGreenandblue.
Sowesetthosetwotomaximumandwezeroouttheportionofred.Andwegetcyancolor.
Color:Magenta
Image:
DecimalCode:
(255,0,255)
Explanation:
Magenta color is formed from the combination of two different colors which are Red and
Blue.Sowesetthosetwotomaximumandwezeroouttheportionofgreen.Andweget
magentacolor.
Color:Yellow
Image:
DecimalCode:
(255,255,0)
Explanation:
52/179
10/26/2016
DIPQuickGuide
YellowcolorisformedfromthecombinationoftwodifferentcolorswhichareRedandGreen.
Sowesetthosetwotomaximumandwezeroouttheportionofblue.Andwegetyellow
color.
Conversion
Nowwewillseethathowcolorareconvertedarefromoneformattoanother.
ConversionfromRGBtoHexcode:
ConversionfromHextorgbisdonethroughthismethod:
Takeacolor.E.g:White=(255,255,255).
Takethefirstportione.g255.
Divideitby16.Likethis:
Takethetwonumbersbelowline,thefactor,andtheremainder.Inthiscaseitis
1515whichisFF.
Repeatthestep2forthenexttwoportions.
Combineallthehexcodeintoone.
Answer:#FFFFFF
ConversionfromHextoRGB:
Conversionfromhexcodetorgbdecimalformatisdoneinthisway.
Takeahexnumber.E.g:#FFFFFF
Breakthisnumberinto3parts:FFFFFF
Takethefirstpartandseparateitscomponents:FF
Converteachofthepartseparatelyintobinary:(1111)(1111)
Nowcombinetheindividualbinariesintoone:11111111
Convertthisbinaryintodecimal:255
53/179
10/26/2016
DIPQuickGuide
Nowrepeatstep2,twomoretimes.
ThevaluecomesinthefirststepisR,secondoneisG,andthethirdonebelongstoB.
Answer:(255,255,255)
CommoncolorsandtheirHexcodehasbeengivenin
thistable.
Color
HexCode
Black
#000000
White
#FFFFFF
Gray
#808080
Red
#FF0000
Green
#00FF00
Blue
#0000FF
Cyan
#00FFFF
Magenta
#FF00FF
Yellow
#FFFF00
GrayscaeltoRGBConversion
WehavealreadydefinetheRGBcolormodelandgrayscaleformatinourtutorialofImagetypes.
Nowwewillconvertancolorimageintoagrayscaleimage.Therearetwomethodstoconvertit.
Bothhastheirownmeritsanddemerits.Themethodsare:
Averagemethod
Weightedmethodorluminositymethod
Averagemethod
Averagemethodisthemostsimpleone.Youjusthavetotaketheaverageofthreecolors.
SinceitsanRGBimage,soitmeansthatyouhaveaddrwithgwithbandthendivideitby
3togetyourdesiredgrayscaleimage.
Itsdoneinthisway.
54/179
10/26/2016
DIPQuickGuide
Grayscale=(R+G+B/3)
Forexample:
If you have an color image like the image shown above and you want to convert it into
grayscaleusingaveragemethod.Thefollowingresultwouldappear.
Explanation
Thereisonethingtobesure,thatsomethinghappenstotheoriginalworks.Itmeansthat
ouraveragemethodworks.Buttheresultswerenotasexpected.Wewantedtoconvertthe
imageintoagrayscale,butthisturnedouttobearatherblackimage.
Problem
Thisproblemariseduetothefact,thatwetakeaverageofthethreecolors.Sincethethree
different colors have three different wavelength and have their own contribution in the
formationofimage,sowehavetotakeaverageaccordingtotheircontribution,notdoneit
averagelyusingaveragemethod.Rightnowwhatwearedoingisthis,
33%ofRed,33%ofGreen,33%ofBlue
Wearetaking33%ofeach,thatmeans,eachoftheportionhassamecontributioninthe
image.Butinrealitythatsnotthecase.Thesolutiontothishasbeengivenbyluminosity
method.
55/179
10/26/2016
DIPQuickGuide
Weightedmethodorluminositymethod
You have seen the problem that occur in the average method. Weighted method has a
solutiontothatproblem.Sinceredcolorhasmorewavelengthofallthethreecolors,and
greenisthecolorthathasnotonlylesswavelengththenredcolorbutalsogreenisthecolor
thatgivesmoresoothingeffecttotheeyes.
It means that we have to decrease the contribution of red color , and increase the
contributionofthegreencolor,andputbluecolorcontributioninbetweenthesetwo.
Sothenewequationthatformis:
Newgrayscaleimage=((0.3*R)+(0.59*G)+(0.11*B)).
Accordingtothisequation,Redhascontribute30%,Greenhascontributed59%whichis
greaterinallthreecolorsandBluehascontributed11%.
Applyingthisequationtotheimage,wegetthis
OriginalImage:
GrayscaleImage:
Explanation
56/179
10/26/2016
DIPQuickGuide
Asyoucanseehere,thattheimagehasnowbeenproperlyconvertedtograyscaleusing
weightedmethod.Ascomparetotheresultofaveragemethod,thisimageismorebrighter.
ConceptofSampling
Conversionofanalogsignaltodigitalsignal:
Theoutputofmostoftheimagesensorsisananalogsignal,andwecannotapplydigital
processingonitbecausewecannotstoreit.Wecannotstoreitbecauseitrequiresinfinite
memorytostoreasignalthatcanhaveinfinitevalues.
Sowehavetoconvertananalogsignalintoadigitalsignal.
To create an image which is digital , we need to covert continuous data into digital form.
Therearetwostepsinwhichitisdone.
Sampling
Quantization
Wewilldiscusssamplingnow,andquantizationwillbediscussedlateronbutfornowonwe
willdiscussjustalittleaboutthedifferencebetweenthesetwoandtheneedofthesetwo
steps.
Basicidea:
Thebasicideabehindconvertingananalogsignaltoitsdigitalsignalis
toconvertbothofitsaxis(x,y)intoadigitalformat.
Sinceanimageiscontinuousnotjustinitscoordinates(xaxis),butalsoinitsamplitude
(yaxis),sothepartthatdealswiththedigitizingofcoordinatesisknownassampling.And
thepartthatdealswithdigitizingtheamplitudeisknownasquantization.
Sampling.
57/179
10/26/2016
DIPQuickGuide
Samplinghasalreadybeenintroducedinourtutorialofintroductiontosignalsandsystem.
Butwearegoingtodiscussheremore.
Herewhatwehavediscussedofthesampling.
Thetermsamplingreferstotakesamples
Wedigitizexaxisinsampling
Itisdoneonindependentvariable
Incaseofequationy=sin(x),itisdoneonxvariable
Itisfurtherdividedintotwoparts,upsamplinganddownsampling
Ifyouwilllookattheabovefigure,youwillseethattherearesomerandomvariationsin
the signal. These variations are due to noise. In sampling we reduce this noise by taking
samples.Itisobviousthatmoresampleswetake,thequalityoftheimagewouldbemore
better,thenoisewouldbemoreremovedandsamehappensviceversa.
However,ifyoutakesamplingonthexaxis,thesignalisnotconvertedtodigitalformat,
unless you take sampling of the yaxis too which is known as quantization. The more
samples eventually means you are collecting more data, and in case of image , it means
morepixels.
Relationshipwithpixels
Sinceapixelisasmallestelementinanimage.Thetotalnumberofpixelsinanimagecan
becalculatedas
Pixels=totalnoofrows*totalnoofcolumns.
Letssaywehavetotalof25pixels,thatmeanswehaveasquareimageof5X5.Thenas
wehavedicussedaboveinsampling,thatmoresampleseventuallyresultinmorepixels.
Soitmeansthatofourcontinuoussignal,wehavetaken25samplesonxaxis.Thatrefers
to25pixelsofthisimage.
58/179
10/26/2016
DIPQuickGuide
ThisleadstoanotherconclusionthatsincepixelisalsothesmallestdivisionofaCCDarray.
SoitmeansithasarelationshipwithCCDarraytoo,whichcanbeexplainedasthis.
RelationshipwithCCDarray
ThenumberofsensorsonaCCDarrayisdirectlyequaltothenumberofpixels.Andsince
we have concluded that the number of pixels is directly equal to the number of samples,
thatmeansthatnumbersampleisdirectlyequaltothenumberofsensorsonCCDarray.
Oversampling.
Inthebeginningwehavedefinethatsamplingisfurthercategorizeintotwotypes.Whichis
upsamplinganddownsampling.Upsamplingisalsocalledasoversampling.
The oversampling has a very deep application in image processing which is known as
Zooming.
Zooming
Wewillformallyintroducezoomingintheupcomingtutorial,butfornowon,wewilljust
brieflyexplainzooming.
Zoomingreferstoincreasethequantityofpixels,sothatwhenyouzoomanimage,you
willseemoredetail.
Theincreaseinthequantityofpixelsisdonethroughoversampling.Theonewaytozoomis
,ortoincreasesamples,istozoomoptically,throughthemotormovementofthelensand
thencapturetheimage.Butwehavetodoit,oncetheimagehasbeencaptured.
Thereisadifferencebetweenzoomingandsampling.
The concept is same , which is, to increase samples. But the key difference is that while
samplingisdoneonthesignals,zoomingisdoneonthedigitalimage.
PixelResolution
Beforewedefinepixelresolution,itisnecessarytodefineapixel.
Pixel
Wehavealreadydefinedapixelinourtutorialofconceptofpixel,inwhichwedefineapixel
as the smallest element of an image. We also defined that a pixel can store a value
proportionaltothelightintensityatthatparticularlocation.
59/179
10/26/2016
DIPQuickGuide
Nowsincewehavedefinedapixel,wearegoingtodefinewhatisresolution.
Resolution
The resolution can be defined in many ways. Such as pixel resolution , spatial resolution ,
temporal resolution , spectral resolution. Out of which we are going to discuss pixel
resolution.
Youhaveprobablyseenthatinyourowncomputersettings,youhavemonitorresolutionof
800x600,640x480e.t.c
Inpixelresolution,thetermresolutionreferstothetotalnumberofcountofpixelsinan
digitalimage.Forexample.IfanimagehasMrowsandNcolumns,thenitsresolutioncan
bedefinedasMXN.
Ifwedefineresolutionasthetotalnumberofpixels,thenpixelresolutioncanbedefined
with set of two numbers. The first number the width of the picture , or the pixels across
columns,andthesecondnumberisheightofthepicture,orthepixelsacrossitswidth.
Wecansaythatthehigheristhepixelresolution,thehigheristhequalityoftheimage.
Wecandefinepixelresolutionofanimageas4500X5500.
Megapixels
Wecancalculatemegapixelsofacamerausingpixelresolution.
Columnpixels(width)Xrowpixels(height)/1Million.
Thesizeofanimagecanbedefinedbyitspixelresolution.
Size=pixelresolutionXbpp(bitsperpixel)
Calculatingthemegapixelsofthecamera
Letssaywehaveanimageofdimension:2500X3192.
Itspixelresolution=2500*3192=7982350bytes.
Dividingitby1million=7.9=8megapixel(approximately).
Aspectratio
Anotherimportantconceptwiththepixelresolutionisaspectratio.
60/179
10/26/2016
DIPQuickGuide
Aspect ratio is the ratio between width of an image and the height of an image. It is
commonly explained as two numbers separated by a colon (8:9). This ratio differs in
differentimages,andindifferentscreens.Thecommonaspectratiosare:
1.33:1,1.37:1,1.43:1,1.50:1,1.56:1,1.66:1,1.75:1,1.78:1,1.85:1,2.00:1,e.t.c
Advantage:
Aspect ratio maintains a balance between the appearance of an image on the screen ,
meansitmaintainsaratiobetweenhorizontalandverticalpixels.Itdoesnotlettheimage
togetdistortedwhenaspectratioisincreased.
Forexample:
This is a sample image , which has 100 rows and 100 columns. If we wish to make is
smaller,andtheconditionisthatthequalityremainsthesameorinotherwaytheimage
doesnotgetdistorted,herehowithappens.
Originalimage:
ChangingtherowsandcolumnsbymaintaintheaspectratioinMSPaint.
Result
Smallerimage,butwithsamebalance.
61/179
10/26/2016
DIPQuickGuide
Youhaveprobablyseenaspectratiosinthevideoplayers,whereyoucanadjustthevideo
accordingtoyourscreenresolution.
Findingthedimensionsoftheimagefromaspectratio:
Aspectratiotellsusmanythings.Withtheaspectratio,youcancalculatethedimensionsof
theimagealongwiththesizeoftheimage.
Forexample
Ifyouaregivenanimagewithaspectratioof6:2ofanimageofpixelresolutionof480000
pixelsgiventheimageisangrayscaleimage.
Andyouareaskedtocalculatetwothings.
Resolvepixelresolutiontocalculatethedimensionsofimage
Calculatethesizeoftheimage
Solution:
Given:
Aspectratio:c:r=6:2
Pixelresolution:c*r=480000
Bitsperpixel:grayscaleimage=8bpp
Find:
Numberofrows=?
Numberofcols=?
Solvingfirstpart:
62/179
10/26/2016
DIPQuickGuide
Solving2ndpart:
Size=rows*cols*bpp
Sizeofimageinbits=400*1200*8=3840000bits
Sizeofimageinbytes=480000bytes
Sizeofimageinkilobytes=48kb(approx).
ConceptofZooming
In this tutorial we are going to introduce the concept of zooming , and the common
techniquesthatareusedtozoomanimage.
Zooming
Zoomingsimplymeansenlargingapictureinasensethatthedetailsintheimagebecame
morevisibleandclear.Zoominganimagehasmanywideapplicationsrangingfromzooming
throughacameralens,tozoomanimageoninternete.t.c.
Forexample
63/179
10/26/2016
DIPQuickGuide
iszoomedinto
Youcanzoomsomethingattwodifferentsteps.
The first step includes zooming before taking an particular image. This is known as pre
processingzoom.Thiszoominvolveshardwareandmechanicalmovement.
The second step is to zoom once an image has been captured. It is done through many
differentalgorithmsinwhichwemanipulatepixelstozoomintherequiredportion.
Wewilldiscussthemindetailinthenexttutorial.
OpticalZoomvsdigitalZoom
Thesetwotypesofzoomaresupportedbythecameras.
OpticalZoom:
The optical zoom is achieved using the movement of the lens of your camera. An optical
zoomisactuallyatruezoom.Theresultoftheopticalzoomisfarbetterthenthatofdigital
zoom.Inopticalzoom,animageismagnifiedbythelensinsuchawaythattheobjectsin
theimageappeartobeclosertothecamera.Inopticalzoomthelensisphysicallyextend
tozoomormagnifyanobject.
DigitalZoom:
64/179
10/26/2016
DIPQuickGuide
Digital zoom is basically image processing within a camera. During a digital zoom , the
centeroftheimageismagnifiedandtheedgesofthepicturegotcropout.Duetomagnified
center,itlookslikethattheobjectisclosertoyou.
During a digital zoom , the pixels got expand , due to which the quality of the image is
compromised.
Thesameeffectofdigitalzoomcanbeseenaftertheimageistakenthroughyourcomputer
byusinganimageprocessingtoolbox/software,suchasPhotoshop.
Thefollowingpictureistheresultofdigitalzoomdonethroughoneofthefollowingmethods
givenbelowinthezoomingmethods.
Nowsinceweareleaningdigitalimageprocessing,wewillnotfocus,onhowanimagecan
be zoomed optically using lens or other stuff. Rather we will focus on the methods, that
enabletozoomadigitalimage.
Zoomingmethods:
Althoughtherearemanymethodsthatdoesthisjob,butwearegoingtodiscussthemost
commonofthemhere.
Theyarelistedbelow.
Pixelreplicationor(Nearestneighborinterpolation)
Zeroorderholdmethod
ZoomingKtimes
Allthesethreemethodsareformallyintroducedinthenexttutorial.
ZoomingMethods
65/179
ZoomingMethods
10/26/2016
DIPQuickGuide
In this tutorial we are going to formally introduce three methods of zooming that were
introducedinthetutorialofIntroductiontozooming.
Methods
Pixelreplicationor(Nearestneighborinterpolation)
Zeroorderholdmethod
ZoomingKtimes
Each of the methods have their own advantages and disadvantages. We will start by
discussingpixelreplication.
Method1:Pixelreplication:
Introduction:
ItisalsoknownasNearestneighborinterpolation.Asitsnamesuggest,inthismethod,we
just replicate the neighboring pixels. As we have already discussed in the tutorial of
Sampling,thatzoomingisnothingbutincreaseamountofsampleorpixels.Thisalgorithm
worksonthesameprinciple.
Working:
Inthismethodwecreatenewpixelsformthealreadygivenpixels.Eachpixelisreplicatedin
thismethodntimesrowwiseandcolumnwiseandyougotazoomedimage.Itsassimple
asthat.
Forexample:
ifyouhaveanimageof2rowsand2columnsandyouwanttozoomittwiceor2times
usingpixelreplication,herehowitcanbedone.
Forabetterunderstanding,theimagehasbeentakenintheformofmatrixwiththepixel
valuesoftheimage.
1
Theaboveimagehastworowsandtwocolumns,wewillfirstzoomitrowwise.
Rowwisezooming:
Whenwezoomitrowwise,wewilljustsimplecopytherowspixelstoitsadjacentnewcell.
66/179
10/26/2016
DIPQuickGuide
Herehowitwouldbedone.
1
Asyoucanthatintheabovematrix,eachpixelisreplicatedtwiceintherows.
Columnsizezooming:
The next step is to replicate each of the pixel column wise, that we will simply copy the
columnpixeltoitsadjacentnewcolumnorsimplybelowit.
Herehowitwouldbedone.
1
Newimagesize:
Asitcanbeseenfromtheaboveexample,thatanoriginalimageof2rowsand2columns
hasbeenconvertedinto4rowsand4columnsafterzooming.Thatmeansthenewimage
hasadimensionsof
(Originalimagerows*zoomingfactor,OriginalImagecols*zoomingfactor)
Advantageanddisadvantage:
Oneoftheadvantageofthiszoomingtechniqueis,itisverysimple.Youjusthavetocopy
thepixelsandnothingelse.
Thedisadvantageofthistechniqueisthatimagegotzoomedbuttheoutputisveryblurry.
And as the zooming factor increased , the image got more and more blurred. That would
eventuallyresultinfullyblurredimage.
Method2:Zeroorderhold
Introduction
Zero order hold method is another method of zooming. It is also known as zoom twice.
Becauseitcanonlyzoomtwice.Wewillseeinthebelowexamplethatwhyitdoesthat.
67/179
10/26/2016
DIPQuickGuide
Working
Inzeroorderholdmethod,wepicktwoadjacentelementsfromtherowsrespectivelyand
thenweaddthemanddividetheresultbytwo,andplacetheirresultinbetweenthosetwo
elements.Wefirstdothisrowwiseandthenwedothiscolumnwise.
Forexample
Letstakeanimageofthedimensionsof2rowsand2columnsandzoomittwiceusingzero
orderhold.
1
Firstwewillzoomitrowwiseandthencolumnwise.
Rowwisezooming
1
Aswetakethefirsttwonumbers:(2+1)=3andthenwedivideitby2,weget1.5which
isapproximatedto1.Thesamemethodisappliedintherow2.
Columnwisezooming
1
Wetaketwoadjacentcolumnpixelvalueswhichare1and3.Weaddthemandgot4.4is
then divided by 2 and we get 2 which is placed in between them. The same method is
appliedinallthecolumns.
Newimagesize
Asyoucanseethatthedimensionsofthenewimageare3x3wheretheoriginalimage
dimensionsare2x2.Soitmeansthatthedimensionsofthenewimagearebasedonthe
followingformula
(2(numberofrows)minus1)X(2(numberofcolumns)minus1)
Advantagesanddisadvantage.
68/179
10/26/2016
DIPQuickGuide
Oneoftheadvantageofthiszoomingtechnique,thatitdoesnotcreateasblurrypictureas
comparetothenearestneighborinterpolationmethod.Butitalsohasadisadvantagethatit
canonlyrunonthepowerof2.Itcanbedemonstratedhere.
Reasonbehindtwicezooming:
Considertheaboveimageof2rowsand2columns.Ifwehavetozoomit6times,using
zeroorderholdmethod,wecannotdoit.Astheformulashowsusthis.
Itcouldonlyzoominthepowerof22,4,8,16,32andsoon.
Evenifyoutrytozoomit,youcannot.Becauseatfirstwhenyouwillzoomittwotimes,
andtheresultwouldbesameasshowninthecolumnwisezoomingwithdimensionsequal
to3x3.Thenyouwillzoomitagainandyouwillgetdimensionsequalto5x5.Nowifyou
willdoitagain,youwillgetdimensionsequalto9x9.
Whereasaccordingtotheformulaofyourstheanswershouldbe11x11.As(6(2)minus1)
X(6(2)minus1)gives11x11.
Method3:KTimeszooming
Introduction:
Ktimesisthethirdzoomingmethodwearegoingtodiscuss.Itisoneofthemostperfect
zoomingalgorithmdiscussedsofar.Itcatersthechallengesofbothtwicezoomingandpixel
replication.Kinthiszoomingalgorithmstandsforzoomingfactor.
Working:
Itworkslikethisway.
Firstofall,youhavetotaketwoadjacentpixelsasyoudidinthezoomingtwice.Thenyou
havetosubtractthesmallerfromthegreaterone.Wecallthisoutput(OP).
Divide the output(OP) with the zooming factor(K). Now you have to add the result to the
smallervalueandputtheresultinbetweenthosetwovalues.
Add the value OP again to the value you just put and place it again next to the previous
puttedvalue.Youhavetodoittillyouplacek1valuesinit.
Repeatthesamestepforalltherowsandthecolumns,andyougetazoomedimages.
Forexample:
Supposeyouhaveanimageof2rowsand3columns,whichisgivenbelow.Andyouhave
tozoomitthriceorthreetimes.
69/179
10/26/2016
DIPQuickGuide
15
30
15
30
15
30
Kinthiscaseis3.K=3.
Thenumberofvaluesthatshouldbeinsertedisk1=31=2.
Rowwisezooming
Takethefirsttwoadjacentpixels.Whichare15and30.
Subtract15from30.3015=15.
Divide15byk.15/k=15/3=5.WecallitOP.(whereopisjustaname)
AddOPtolowernumber.15+OP=15+5=20.
AddOPto20again.20+OP=20+5=25.
Wedothat2timesbecausewehavetoinsertk1values.
Nowrepeatthisstepforthenexttwoadjacentpixels.Itisshowninthefirsttable.
Afterinsertingthevalues,youhavetosorttheinsertedvaluesinascendingorder,sothere
remainsasymmetrybetweenthem.
Itisshowninthesecondtable
Table1.
15
20
25
30
20
25
15
30
20
25
15
20
25
30
Table2.
Columnwisezooming
Thesameprocedurehastobeperformedcolumnwise.Theprocedureincludetakingthetwo
adjacentpixelvalues,andthensubtractingthesmallerfromthebiggerone.Thenafterthat
,youhavetodivideitbyk.StoretheresultasOP.AddOPtosmallerone,andthenagain
addOPtothevaluethatcomesinfirstadditionofOP.Insertthenewvalues.
70/179
10/26/2016
DIPQuickGuide
Herewhatyougotafterallthat.
15
20
25
30
25
20
15
20
21
21
25
21
21
20
25
22
22
20
22
22
25
30
25
20
15
20
25
30
Newimagesize
Thebestwaytocalculatetheformulaforthedimensionsofanewimageistocomparethe
dimensionsoftheoriginalimageandthefinalimage.Thedimensionsoftheoriginalimage
were2X3.Andthedimensionsofthenewimageare4x7.
Theformulathusis:
(K(numberofrowsminus1)+1)X(K(numberofcolsminus1)+1)
Advantagesanddisadvantages
Theoneoftheclearadvantagethatktimezoomingalgorithmhasthatitisabletocompute
zoom of any factor which was the power of pixel replication algorithm , also it gives
improvedresult(lessblurry)whichwasthepowerofzeroorderholdmethod.SohenceIt
comprisesthepowerofthetwoalgorithms.
Theonlydifficultythisalgorithmhasthatithastobesortintheend,whichisanadditional
step,andthusincreasesthecostofcomputation.
SpatialResolution
Imageresolution
Imageresolutioncanbedefinedinmanyways.Onetypeofitwhichispixelresolutionthat
hasbeendiscussedinthetutorialofpixelresolutionandaspectratio.
Inthistutorial,wearegoingtodefineanothertypeofresolutionwhichisspatialresolution.
Spatialresolution:
Spatial resolution states that the clarity of an image cannot be determined by the pixel
resolution.Thenumberofpixelsinanimagedoesnotmatter.
Spatialresolutioncanbedefinedasthe
71/179
10/26/2016
DIPQuickGuide
smallestdiscernibledetailinanimage.(DigitalImageProcessingGonzalez,Woods2nd
Edition)
Orinotherwaywecandefinespatialresolutionasthenumberofindependentpixelsvalues
perinch.
Inshortwhatspatialresolutionreferstoisthatwecannotcomparetwodifferenttypesof
imagestoseethatwhichoneisclearorwhichoneisnot.Ifwehavetocomparethetwo
images,toseewhichoneismoreclearorwhichhasmorespatialresolution,wehaveto
comparetwoimagesofthesamesize.
Forexample:
Youcannotcomparethesetwoimagestoseetheclarityoftheimage.
Althoughbothimagesareofthesameperson,butthatisnottheconditionwearejudging
on.ThepictureontheleftiszoomedoutpictureofEinsteinwithdimensionsof227x222.
Whereas the picture on the right side has the dimensions of 980 X 749 and also it is a
zoomedimage.Wecannotcomparethemtoseethatwhichoneismoreclear.Remember
the factor of zoom does not matter in this condition, the only thing that matters is that
thesetwopicturesarenotequal.
Soinordertomeasurespatialresolution,thepicturesbelowwouldserverthepurpose.
72/179
10/26/2016
DIPQuickGuide
Nowyoucancomparethesetwopictures.Boththepictureshassamedimensionswhichare
of227X222.Nowwhenyoucomparethem,youwillseethatthepictureontheleftside
has more spatial resolution or it is more clear then the picture on the right side. That is
becausethepictureontherightisablurredimage.
Measuringspatialresolution:
Sincethespatialresolutionreferstoclarity,sofordifferentdevices,differentmeasurehas
beenmadetomeasureit.
Forexample:
Dotsperinch
Linesperinch
Pixelsperinch
Theyarediscussedinmoredetailinthenexttutorialbutjustabriefintroductionhasbeen
givenbelow.
Dotsperinch:
DotsperinchorDPIisusuallyusedinmonitors.
Linesperinch:
LinesperinchorLPIisusuallyusedinlaserprinters.
Pixelperinch:
73/179
10/26/2016
DIPQuickGuide
PixelperinchorPPIismeasurefordifferentdevicessuchastablets,Mobilephonese.t.c.
Pixels,DotsandLinesPerInch
Intheprevioustutorialofspatialresolution,wediscussedthebriefintroductionofPPI,DPI,
LPI.Nowweareformallygoingtodiscussallofthem.
Pixelsperinch.
PixeldensityorPixelsperinchisameasureofspatialresolutionfordifferentdevicesthat
includestablets,mobilephones.
ThehigheristhePPI,thehigheristhequality.Inordertomoreunderstandit,thathowit
calculated.LetscalculatethePPIofamobilephone.
Calculatingpixelsperinch(PPI)ofSamsunggalaxyS4:
TheSamsunggalaxys4hasPPIorpixeldensityof441.Buthowdoesitiscalculated?
FirstofallwewillPythagorastheoremtocalculatethediagonalresolutioninpixels.
Itcanbegivenas:
Whereaandbaretheheightandwidthresolutionsinpixelandcisthediagonalresolution
inpixels.
ForSamsunggalaxys4,itis1080x1920pixels.
Soputtingthosevaluesintheequationgivestheresult
C=2202.90717
74/179
10/26/2016
DIPQuickGuide
NowwewillcalculatePPI
PPI=c/diagonalsizeininches
ThediagonalsizeininchesofSamsungalaxys4is5.0inches,whichcanbeconfirmedfrom
anywhere.
PPI=2202.90717/5.0
PPI=440.58
PPI=441(approx)
ThatmeansthatthepixeldensityofSamsunggalaxys4is441PPI.
Dotsperinch.
ThedpiisoftenrelatetoPPI,whereasthereisadifferencebetweenthesetwo.DPIordots
perinchisameasureofspatialresolutionofprinters.Incaseofprinters,dpimeansthat
howmanydotsofinkareprintedperinchwhenanimagegetprintedoutfromtheprinter.
Remember,itisnotnecessarythateachPixelperinchisprintedbyonedotperinch.There
maybemanydotsperinchusedforprintingonepixel.Thereasonbehindthisthatmostof
thecolorprintersusesCMYKmodel.Thecolorsarelimited.Printerhastochoosefromthese
colorstomakethecolorofthepixelwhereaswithinpc,youhavehundredsofthousandsof
colors.
The higher is the dpi of the printer , the higher is the quality of the printed document or
imageonpaper.
Usuallysomeofthelaserprintershavedpiof300andsomehave600ormore.
Linesperinch.
Whendpireferstodotsperinch,linerperinchreferstolinesofdotsperinch.Theresolution
ofhalftonescreenismeasuredinlinesperinch.
Thefollowingtableshowssomeofthelinesperinchcapacityoftheprinters.
Printer
LPI
Screenprinting
4565lpi
Laserprinter(300
dpi)
65lpi
75/179
10/26/2016
DIPQuickGuide
Laserprinter(600
dpi)
85105lpi
OffsetPress
(newsprintpaper)
85lpi
OffsetPress(coated
paper)
85185lpi
GrayLevelResolution
Imageresolution:
Aresolutioncanbedefinedasthetotalnumberofpixelsinanimage.Thishasbeendiscussedin
Imageresolution.Andwehavealsodiscussed,thatclarityofanimagedoesnotdependson
numberofpixels,butonthespatialresolutionoftheimage.Thishasbeendiscussedinthespatial
resolution.Herewearegoingtodiscussanothertypeofresolutionwhichiscalledgraylevel
resolution.
Graylevelresolution:
Graylevelresolutionreferstothepredictableordeterministicchangeintheshadesorlevels
ofgrayinanimage.
Inshortgraylevelresolutionisequaltothenumberofbitsperpixel.
Wehavealreadydiscussedbitsperpixelinourtutorialofbitsperpixelandimagestorage
requirements.Wewilldefinebppherebriefly.
BPP:
Thenumberofdifferentcolorsinanimageisdependsonthedepthofcolororbitsperpixel.
Mathematically:
Themathematicalrelationthatcanbeestablishedbetweengraylevelresolutionandbitsper
pixelcanbegivenas.
InthisequationLreferstonumberofgraylevels.Itcanalsobedefinedastheshadesof
gray.Andkreferstobpporbitsperpixel.Sothe2raisetothepowerofbitsperpixelis
equaltothegraylevelresolution.
76/179
10/26/2016
DIPQuickGuide
Forexample:
TheaboveimageofEinstein is an gray scale image. Means it is an image with 8 bits per

pixelor8bpp.
Nowifweretocalculatethegraylevelresolution,herehowwegonnadoit.
Itmeansitgraylevelresolutionis256.Orinotherwaywecansaythatthisimagehas256
differentshadesofgray.
Themoreisthebitsperpixelofanimage,themoreisitsgraylevelresolution.
Defininggraylevelresolutionintermsofbpp:
Itisnotnecessarythatagraylevelresolutionshouldonlybedefinedintermsoflevels.We
canalsodefineitintermsofbitsperpixel.
Forexample:
Ifyouaregivenanimageof4bpp,andyouareaskedtocalculateitsgraylevelresolution.
Therearetwoanswerstothatquestion.
Thefirstansweris16levels.
Thesecondansweris4bits.
77/179
10/26/2016
DIPQuickGuide
FindingbppfromGraylevelresolution:
Youcanalsofindthebitsperpixelsfromthegivengraylevelresolution.Forthis,wejust
havetotwisttheformulaalittle.
Equation1.
Thisformulafindsthelevels.Nowifweweretofindthebitsperpixelorinthiscasek,we
willsimplychangeitlikethis.
K=logbase2(L)Equation(2)
Because in the first equation the relationship between Levels (L ) and bits per pixel (k) is
exponentional.Nowwehavetorevertit,andthustheinverseofexponentionalislog.
Letstakeanexampletofindbitsperpixelfromgraylevelresolution.
Forexample:
Ifyouaregivenanimageof256levels.Whatisthebitsperpixelrequiredforit.
Putting256intheequation,weget.
K=logbase2(256)
K=8.
Sotheansweris8bitsperpixel.
Graylevelresolutionandquantization:
Thequantizationwillbeformallyintroducedinthenexttutorial,butherewearejustgoing
toexplaintherelationshipbetweengraylevelresolutionandquantization.
Gray level resolution is found on the y axis of the signal. In the tutorial of Introduction to
signalsandsystem,wehavestudiedthatdigitizingaananalogsignalrequirestwosteps.
Samplingandquantization.
78/179
10/26/2016
DIPQuickGuide
Samplingisdoneonxaxis.AndquantizationisdoneinYaxis.
Sothatmeansdigitizingthegraylevelresolutionofanimageisdoneinquantization.
ConceptofQuantization
Wehaveintroducedquantizationinourtutorialofsignalsandsystem.Weareformallygoing
torelateitwithdigitalimagesinthistutorial.Letsdiscussfirstalittlebitaboutquantization.
Digitizingasignal.
As we have seen in the previous tutorials , that digitizing an analog signal into a digital ,
requires two basic steps. Sampling and quantization. Sampling is done on x axis. It is the
conversionofxaxis(infinitevalues)todigitalvalues.
Thebelowfigureshowssamplingofasignal.
Samplingwithrelationtodigitalimages:
The concept of sampling is directly related to zooming. The more samples you take , the
morepixels,youget.Oversamplingcanalsobecalledaszooming.Thishasbeendiscussed
undersamplingandzoomingtutorial.
But the story of digitizing a signal does not end at sampling too, there is another step
involvedwhichisknownasQuantization.
79/179
10/26/2016
DIPQuickGuide
Whatisquantization.
Quantization is opposite to sampling. It is done on y axis. When you are qunaitizing an
image,youareactuallydividingasignalintoquanta(partitions).
On the x axis of the signal , are the coordinate values, and on the y axis , we have
amplitudes.SodigitizingtheamplitudesisknownasQuantization.
Herehowitisdone
You can see in this image , that the signal has been quantified into three different levels.
That means that when we sample an image , we actually gather a lot of values, and in
quantization,wesetlevelstothesevalues.Thiscanbemoreclearintheimagebelow.
Inthefigureshowninsampling,althoughthesampleshasbeentaken,buttheywerestill
spanningverticallytoacontinuousrangeofgraylevelvalues.Inthefigureshownabove,
80/179
10/26/2016
DIPQuickGuide
these vertically ranging values have been quantized into 5 different levels or partitions.
Rangingfrom0blackto4white.Thislevelcouldvaryaccordingtothetypeofimageyou
want.
Therelationofquantizationwithgraylevelshasbeenfurtherdiscussedbelow.
RelationofQuantizationwithgraylevelresolution:
The quantized figure shown above has 5 different levels of gray. It means that the image
formedfromthissignal,wouldonlyhave5differentcolors.Itwouldbeablackandwhite
imagemoreorlesswithsomecolorsofgray.Nowifyouweretomakethequalityofthe
imagemorebetter,thereisonethingyoucandohere.Whichis,toincreasethelevels,or
graylevelresolutionup.Ifyouincreasethislevelto256,itmeansyouhaveangrayscale
image.Whichisfarbetterthensimpleblackandwhiteimage.
Now256,or5orwhateverlevelyouchooseiscalledgraylevel.Remembertheformula
thatwediscussedintheprevioustutorialofgraylevelresolutionwhichis,
Wehavediscussedthatgraylevelcanbedefinedintwoways.Whichwerethesetwo.
Graylevel=numberofbitsperpixel(BPP).(kintheequation)
Graylevel=numberoflevelsperpixel.
Inthiscasewehavegraylevelisequalto256.Ifwehavetocalculatethenumberofbits,
wewouldsimplyputthevaluesintheequation.Incaseof256levels,wehave256different
shadesofgrayand8bitsperpixel,hencetheimagewouldbeagrayscaleimage.
Reducingthegraylevel
Nowwewillreducethegraylevelsoftheimagetoseetheeffectontheimage.
Forexample:
Letssayyouhaveanimageof8bpp,thathas256differentlevels.Itisagrayscaleimage
andtheimagelookssomethinglikethis.
256GrayLevels
81/179
10/26/2016
DIPQuickGuide
Nowwewillstartreducingthegraylevels.Wewillfirstreducethegraylevelsfrom256to
128.
128GrayLevels
There is not much effect on an image after decrease the gray levels to its half. Lets
decreasesomemore.
64GrayLevels
Stillnotmucheffect,thenletsreducethelevelsmore.
32GrayLevels
82/179
10/26/2016
DIPQuickGuide
Surprisedtosee,thatthereisstillsomelittleeffect.Maybeitsduetoreason,thatitis
thepictureofEinstein,butletsreducethelevelsmore.
16GrayLevels
Boomhere,wego,theimagefinallyreveals,thatitiseffectedbythelevels.
8GrayLevels
4GrayLevels
83/179
10/26/2016
DIPQuickGuide
Nowbeforereducingit,furthertwo2levels,youcaneasilyseethattheimagehasbeen
distorted badly by reducing the gray levels. Now we will reduce it to 2 levels, which is
nothingbutasimpleblackandwhitelevel.Itmeanstheimagewouldbesimpleblackand
whiteimage.
2GrayLevels
Thatsthelastlevelwecanachieve,becauseifreduceitfurther,itwouldbesimplyablack
image,whichcannotbeinterpreted.
Contouring:
There is an interesting observation here , that as we reduce the number of gray levels ,
thereisaspecialtypeofeffectstartappearingintheimage,whichcanbeseenclearin16
graylevelpicture.ThiseffectisknownasContouring.
Isopreferencecurves:
The answer to this effect , that why it appears , lies in Iso preference curves. They are
discussedinournexttutorialofContouringandIsopreferencecurves.
ISOpreferencecurves
Whatiscontouring?
Aswedecreasethenumberofgraylevelsinanimage,somefalsecolors,oredgesstart
appearingonanimage.ThishasbeenshowninourlasttutorialofQuantization.
Letshavealookatit.
Considerwe,haveanimageof8bpp(agrayscaleimage)with256differentshadesofgray
orgraylevels.
84/179
10/26/2016
DIPQuickGuide
This above picture has 256 different shades of gray. Now when we reduce it to 128 and
furtherreduceit64,theimageismoreorlessthesame.Butwhenrereduceitfurtherto32
differentlevels,wegotapicturelikethis
If you will look closely , you will find that the effects start appearing on the image.These
effects are more visible when we reduce it further to 16 levels and we got an image like
this.
Theselines,thatstartappearingonthisimageareknownascontouringthatareverymuch
visibleintheaboveimage.
Increaseanddecreaseincontouring
The effect of contouring increase as we reduce the number of gray levels and the effect
decreaseasweincreasethenumberofgraylevels.Theyarebothviceversa
85/179
10/26/2016
DIPQuickGuide
VS
That means more quantization , will effect in more contouring and vice versa. But is this
alwaysthecase.TheanswerisNo.Thatdependsonsomethingelsethatisdiscussedbelow.
Isopreferencecurves
Astudyconductedonthiseffectofgraylevelandcontouring,andtheresultswereshownin
thegraphintheformofcurves,knownasIsopreferencecurves.
The phenomena of Isopreference curves shows , that the effect of contouring not only
dependsonthedecreasingofgraylevelresolutionbutalsoontheimagedetail.
Theessenceofthestudyis:
Ifanimagehasmoredetail,theeffectofcontouringwouldstartappearonthisimagelater,
ascomparetoanimagewhichhaslessdetail,whenthegraylevelsarequantized.
Accordingtotheoriginalresearch,theresearcherstookthesethreeimagesandtheyvary
theGraylevelresolution,inallthreeimages.
Theimageswere
86/179
10/26/2016
DIPQuickGuide
Levelofdetail:
Thefirstimagehasonlyafaceinit,andhenceverylessdetail.Thesecondimagehassome
other objects in the image too , such as camera man, his camera , camera stand, and
background objects e.t.c. Whereas the third image has more details then all the other
images.
Experiment:
Thegraylevelresolutionwasvariedinalltheimages,andtheaudiencewasaskedtorate
these three images subjectively. After the rating , a graph was drawn according to the
results.
87/179
10/26/2016
DIPQuickGuide
Result:
The result was drawn on the graph. Each curve on the graph represents one image. The
values on the x axis represents the number of gray levels and the values on the y axis
representsbitsperpixel(k).
Thegraphhasbeenshownbelow.
Accordingtothisgraph,wecanseethatthefirstimagewhichwasofface,wassubjectto
contouring early then all of the other two images. The second image , that was of the
cameraman was subject to contouring a bit after the first image when its gray levels are
reduced.Thisisbecauseithasmoredetailsthenthefirstimage.Andthethirdimagewas
subjecttocontouringalotafterthefirsttwoimagesie:after4bpp.Thisisbecause,this
imagehasmoredetails.
Conclusion:
Soformoredetailedimages,theisopreferencecurvesbecomemoreandmorevertical.It
also means that for an image with a large amount of details , very few gray levels are
needed.
ConceptofDithering
In the last two tutorials of Quantization and contouring , we have seen that reducing the
gray level of an image reduces the number of colors required to denote an image. If the
88/179
10/26/2016
DIPQuickGuide
gray levels are reduced two 2 , the image that appears doesnot have much spatial
resolutionorisnotverymuchappealing.
Dithering:
Dithering is the process by which we create illusions of the color that are not present
actually.Itisdonebytherandomarrangementofpixels.
Forexample.Considerthisimage.
Thisisanimagewithonlyblackandwhitepixelsinit.Itspixelsarearrangedinanorderto
form another image that is shown below. Note at the arrangement of pixels has been
changed,butnotthequantityofpixels.
WhyDithering?
Whydoweneeddithering,theanswerofthisliesinitsrelationwithquantization.
Ditheringwithquantization.
Whenweperformquantization,tothelastlevel,weseethattheimagethatcomesinthe
lastlevel(level2)lookslikethis.
Nowaswecanseefromtheimagehere,thatthepictureisnotveryclear,especiallyifyou
will look at the left arm and back of the image of the Einstein. Also this picture does not
havemuchinformationordetailoftheEinstein.
Nowifweweretochangethisimageintosomeimagethatgivesmoredetailthenthis,we
havetoperformdithering.
89/179
10/26/2016
DIPQuickGuide
Performingdithering.
First of all , we will work on threholding. Dithering is usually working to improve
thresholding.Duringthreholding,thesharpedgesappearwheregradientsaresmoothinan
image.
In thresholding , we simply choose a constant value. All the pixels above that value are
consideredas1andallthevaluebelowitareconsideredas0.
Wegotthisimageafterthresholding.
Sincethereisnotmuchchangeintheimage,asthevaluesarealready0and1orblack
andwhiteinthisimage.
Nowweperformsomerandomditheringtoit.Itssomerandomarrangementofpixels.
Wegotanimagethatgivesslighterofthemoredetails,butitscontrastisverylow.
Sowedosomemoreditheringthatwillincreasethecontrast.Theimagethatwegotisthis:
90/179
10/26/2016
DIPQuickGuide
Nowwemixtheconceptsofrandomdithering,alongwiththresholdandwegotanimage
likethis.
Nowyousee,wegotalltheseimagesbyjustrearrangingthepixelsofanimage.Thisre
arrangingcouldberandomorcouldbeaccordingtosomemeasure.
HistogramsIntroduction
Before discussing the use of Histograms in image processing , we will first look at what
histogramis,howitisusedandthenanexampleofhistogramstohavemoreunderstanding
ofhistogram.
Histograms:
Ahistogramisagraph.Agraphthatshowsfrequencyofanything.Usuallyhistogramhave
barsthatrepresentfrequencyofoccurringofdatainthewholedataset.
AHistogramhastwoaxisthexaxisandtheyaxis.
Thexaxiscontainseventwhosefrequencyyouhavetocount.
Theyaxiscontainsfrequency.
Thedifferentheightsofbarshowsdifferentfrequencyofoccurrenceofdata.
Usuallyahistogramlookslikethis.
91/179
10/26/2016
DIPQuickGuide
Nowwewillseeanexampleofthishistogramisbuild
Example:
Consideraclassofprogrammingstudentsandyouareteachingpythontothem.
Attheendofthesemester,yougotthisresultthatisshownintable.Butitisverymessy
and doesnot show your overall result of class. So you have to make a histogram of your
result,showingtheoverallfrequencyofoccurrenceofgradesinyourclass.Herehowyou
aregoingtodoit.
Resultsheet:
Name
Grade
John
Jack
Carter
Tommy
Lisa
C+
Derek
Tom
B+
Histogramofresultsheet:
Nowwhatyouaregoingtodois,thatyouhavetofindwhatcomesonthexandtheyaxis.
92/179
10/26/2016
DIPQuickGuide
Thereisonethingtobesure,thatyaxiscontainsthefrequency,sowhatcomesonthex
axis. X axis contains the event whose frequency has to be calculated. In this case x axis
containsgrades.
Nowwewillhowdoweuseahistograminanimage.
Histogramofanimage
Histogram of an image , like other histograms also shows frequency. But an image
histogram , shows frequency of pixels intensity values. In an image histogram, the x axis
showsthegraylevelintensitiesandtheyaxisshowsthefrequencyoftheseintensities.
Forexample:
ThehistogramoftheabovepictureoftheEinsteinwouldbesomethinglikethis
93/179
10/26/2016
DIPQuickGuide
Thexaxisofthehistogramshowstherangeofpixelvalues.Sinceitsan8bppimage,that
meansithas256levelsofgrayorshadesofgrayinit.Thatswhytherangeofxaxisstarts
from 0 and end at 255 with a gap of 50. Whereas on the y axis , is the count of these
intensities.
Asyoucanseefromthegraph,thatmostofthebarsthathavehighfrequencyliesinthe
first half portion which is the darker portion. That means that the image we have got is
darker.Andthiscanbeprovedfromtheimagetoo.
ApplicationsofHistograms:
Histogramshasmanyusesinimageprocessing.Thefirstuseasithasalsobeendiscussed
above is the analysis of the image. We can predict about an image by just looking at its
histogram.Itslikelookinganxrayofaboneofabody.
Theseconduseofhistogramisforbrightnesspurposes.Thehistogramshaswideapplication
in image brightness. Not only in brightness , but histograms are also used in adjusting
contrastofanimage.
Anotherimportantuseofhistogramistoequalizeanimage.
And last but not the least, histogram has wide use in thresholding. This is mostly used in
computervision.
BrightnessandContrast
94/179
10/26/2016
DIPQuickGuide
Brightness:
Brightness is a relative term. It depends on your visual perception. Since brightness is a
relativeterm,sobrightnesscanbedefinedastheamountofenergyoutputbyasourceof
lightrelativetothesourcewearecomparingitto.Insomecaseswecaneasilysaythatthe
imageisbright,andinsomecases,itsnoteasytoperceive.
Forexample:
Justhavealookatbothoftheseimages,andcomparewhichoneisbrighter.
Wecaneasilysee,thattheimageontherightsideisbrighterascomparedtotheimageon
theleft.
Butiftheimageontherightismademoredarkerthenthefirstone,thenwecansaythat
theimageontheleftismorebrighterthentheleft.
Howtomakeanimagebrighter.
Brightness can be simply increased or decreased by simple addition or subtraction, to the
imagematrix.
Considerthisblackimageof5rowsand5columns
95/179
10/26/2016
DIPQuickGuide
Sincewealreadyknow,thateachimagehasamatrixatitsbehindthatcontainsthepixel
values.Thisimagematrixisgivenbelow.
0
Sincethewholematrixisfilledwithzero,andtheimageisverymuchdarker.
Nowwewillcompareitwithanothersameblackimagetoseethisimagegotbrighterornot.
Stillboththeimagesaresame,nowwewillperformsomeoperationsonimage1,dueto
whichitbecomesbrighterthenthesecondone.
Whatwewilldois,thatwewillsimplyaddavalueof1toeachofthematrixvalueofimage
1.Afteraddingtheimage1wouldsomethinglikethis.
96/179
10/26/2016
DIPQuickGuide
Nowwewillagaincompareitwithimage2,andseeanydifference.
Wesee,thatstillwecannottellwhichimageisbrighterasbothimageslooksthesame.
Nowwhatwewilldo,isthatwewilladd50toeachofthematrixvalueoftheimage1and
seewhattheimagehasbecome.
Theoutputisgivenbelow.
Nowagain,wewillcompareitwithimage2.
Nowyoucanseethattheimage1isslightlybrighterthentheimage2.Wegoon,andadd
another45valuetoitsmatrixofimage1,andthistimewecompareagainbothimages.
97/179
10/26/2016
DIPQuickGuide
Nowwhenyoucompareit,youcanseethatthisimage1isclearlybrighterthentheimage
2.
Evenitisbrighterthentheoldimage1.Atthispointthematrixoftheimage1contains100
ateachindexasfirstadd5,then50,then45.So5+50+45=100.
Contrast
Contrastcanbesimplyexplainedasthedifferencebetweenmaximumandminimumpixel
intensityinanimage.
Forexample.
Considerthefinalimage1inbrightness.
Thematrixofthisimageis:
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
98/179
10/26/2016
DIPQuickGuide
Themaximumvalueinthismatrixis100.
Theminimumvalueinthismatrixis100.
Contrast=maximumpixelintensity(subtractedby)minimumpixelintensity
=100(subtractedby)100
=0
0meansthatthisimagehas0contrast.
ImageTransformations
Beforewediscuss,whatisimagetransformation,wewilldiscusswhatatransformationis.
Transformation.
Transformationisafunction.Afunctionthatmapsonesettoanothersetafterperforming
someoperations.
DigitalImageProcessingsystem:
Wehavealreadyseenintheintroductorytutorialsthatindigitalimageprocessing,wewill
developasystemthatwhoseinputwouldbeanimageandoutputwouldbeanimagetoo.
Andthesystemwouldperformsomeprocessingontheinputimageandgivesitsoutputas
anprocessedimage.Itisshownbelow.
Now function applied inside this digital system that process an image and convert it into
outputcanbecalledastransformationfunction.
99/179
10/26/2016
DIPQuickGuide
Asitshowstransformationorrelation,thathowanimage1isconvertedtoimage2.
Imagetransformation.
Considerthisequation
G(x,y)=T{f(x,y)}
Inthisequation,
F(x,y)=inputimageonwhichtransformationfunctionhastobeapplied.
G(x,y)=theoutputimageorprocessedimage.
Tisthetransformationfunction.
Thisrelationbetweeninputimageandtheprocessedoutputimagecanalsoberepresented
as.
s=T(r)
whererisactuallythepixelvalueorgraylevelintensityoff(x,y)atanypoint.Andsisthe
pixelvalueorgraylevelintensityofg(x,y)atanypoint.
The basic gray level transformation has been discussed in our tutorial of basic gray level
transformations.
Nowwearegoingtodiscusssomeoftheverybasictransformationfunctions.
Examples:
Considerthistransformationfunction.
Letstakethepointrtobe256,andthepointptobe127.Considerthisimagetobeaone
bppimage.Thatmeanswehaveonlytwolevelsofintensitiesthatare0and1.Sointhis
100/179
10/26/2016
DIPQuickGuide
casethetransformationshownbythegraphcanbeexplainedas.
Allthepixelintensityvaluesthatarebelow127(pointp)are0,meansblack.Andallthe
pixelintensityvaluesthataregreaterthen127,are1,thatmeanswhite.Butattheexact
pointof127,thereisasuddenchangeintransmission,sowecannottellthatatthatexact
point,thevaluewouldbe0or1.
Mathematicallythistransformationfunctioncanbedenotedas:
Consideranothertransformationlikethis:
Nowifyouwilllookatthisparticulargraph,youwillseeastraighttransitionlinebetween
inputimageandoutputimage.
Itshowsthatforeachpixelorintensityvalueofinputimage,thereisasameintensityvalue
ofoutputimage.Thatmeanstheoutputimageisexactreplicaoftheinputimage.
Itcanbemathematicallyrepresentedas:
g(x,y)=f(x,y)
theinputandoutputimagewouldbeinthiscaseareshownbelow.
101/179
10/26/2016
DIPQuickGuide
HistogramSliding
The basic concept of histograms has been discussed in the tutorial of Introduction to
histograms.Butwewillbrieflyintroducethehistogramhere.
Histogram:
Histogram is nothing but a graph that shows frequency of occurrence of data. Histograms
hasmanyuseinimageprocessing,outofwhichwearegoingtodiscussoneuserherewhich
iscalledhistogramsliding.
Histogramsliding.
Inhistogramsliding,wejustsimplyshiftacompletehistogramrightwardsorleftwards.Due
toshiftingorslidingofhistogramtowardsrightorleft,aclearchangecanbeseeninthe
image.Inthistutorialwearegoingtousehistogramslidingformanipulatingbrightness.
Thetermie:Brightnesshasbeendiscussedinourtutorialofintroductiontobrightnessand
contrast.Butwearegoingtobrieflydefinehere.
Brightness:
Brightness is a relative term. Brightness can be defined as intensity of light emit by a
particularlightsource.
Contrast:
Contrastcanbedefinedasthedifferencebetweenmaximumandminimumpixelintensityin
animage.
SlidingHistograms
Increasingbrightnessusinghistogramsliding
102/179
10/26/2016
DIPQuickGuide
Histogramofthisimagehasbeenshownbelow.
Ontheyaxisofthishistogramarethefrequencyorcount.Andonthexaxis,wehavegray
level values. As you can see from the above histogram , that those gray level intensities
whosecountismorethen700,liesinthefirsthalfportion,meanstowardsblackerportion.
Thatswhywegotanimagethatisabitdarker.
Inordertobrightit,wewillslideitshistogramtowardsright,ortowardswhiterportion.In
ordertodoweneedtoaddatleastavalueof50tothisimage.Becausewecanseefrom
thehistogramabove,thatthisimagealsohas0pixelintensities,thatarepureblack.Soif
weadd0to50,wewillshiftallthevaluesliesat0intensityto50intensityandalltherest
ofthevalueswillbeshiftedaccordingly.
Letsdoit.
Herewhatwegotafteradding50toeachpixelintensity.
Theimagehasbeenshownbelow.
103/179
10/26/2016
DIPQuickGuide
Anditshistogramhasbeenshownbelow.
Letscomparethesetwoimagesandtheirhistogramstoseethatwhatchangehavetogot.
Conclusion:
As we can clearly see from the new histogram that all the pixels values has been shifted
towardsrightanditseffectcanbeseeninthenewimage.
104/179
10/26/2016
DIPQuickGuide
Decreasingbrightnessusinghistogramsliding
Now if we were to decrease brightness of this new image to such an extent that the old
imagelookbrighter,wegottosubtractsomevaluefromallthematrixofthenewimage.
Thevaluewhichwearegoingtosubtractis80.Becausewealreadyadd50totheoriginal
image and we got a new brighter image, now if we want to make it darker , we have to
subtractatleastmorethan50fromit.
Andthiswhatwegotaftersubtracting80fromthenewimage.
Conclusion:
Itisclearfromthehistogramofthenewimage,thatallthepixelvalueshasbeenshifted
towardsrightandthus,itcanbevalidatedfromtheimagethatnewimageisdarkerand
nowtheoriginalimagelookbrighterascomparetothisnewimage.
Histogramstretching
OneoftheotheradvantageofHistogramsthatwediscussedinourtutorialofintroduction
tohistogramsiscontrastenhancement.
There are two methods of enhancing contrast. The first one is called Histogram stretching
that increase contrast. The second one is called Histogram equalization that enhance
contrastandithasbeendiscussedinourtutorialofhistogramequalization.
105/179
10/26/2016
DIPQuickGuide
Beforewewilldiscussthehistogramstretchingtoincreasecontrast,wewillbrieflydefine
contrast.
Contrast.
Contrastisthedifferencebetweenmaximumandminimumpixelintensity.
Considerthisimage.
Thehistogramofthisimageisshownbelow.
Nowwecalculatecontrastfromthisimage.
Contrast=225.
106/179
10/26/2016
DIPQuickGuide
Nowwewillincreasethecontrastoftheimage.
Increasingthecontrastoftheimage:
Theformulaforstretchingthehistogramoftheimagetoincreasethecontrastis
Theformularequiresfindingtheminimumandmaximumpixelintensitymultiplybylevelsof
gray.Inourcasetheimageis8bpp,solevelsofgrayare256.
Theminimumvalueis0andthemaximumvalueis225.Sotheformulainourcaseis
wheref(x,y)denotesthevalueofeachpixelintensity.Foreachf(x,y)inanimage,wewill
calculatethisformula.
Afterdoingthis,wewillbeabletoenhanceourcontrast.
Thefollowingimageappearafterapplyinghistogramstretching.
Thestretchedhistogramofthisimagehasbeenshownbelow.
Note the shape and symmetry of histogram. The histogram is now stretched or in other
meansexpand.Havealookatit.
107/179
10/26/2016
DIPQuickGuide
Inthiscasethecontrastoftheimagecanbecalculatedas
Contrast=240
Hencewecansaythatthecontrastoftheimageisincreased.
Note:thismethodofincreasingcontrastdoesnotworkalways,butitfailsonsomecases.
Failingofhistogramstretching
Aswehavediscussed,thatthealgorithmfailsonsomecases.Thosecasesincludeimages
withwhenthereispixelintensity0and255arepresentintheimage
Becausewhenpixelintensities0and255arepresentinanimage,theninthatcasethey
becometheminimumandmaximumpixelintensitywhichruinstheformulalikethis.
OriginalFormula
Puttingfailcasevaluesintheformula:
108/179
10/26/2016
DIPQuickGuide
Simplifythatexpressiongives
That means the output image is equal to the processed image. That means there is no
effectofhistogramstretchinghasbeendoneatthisimage.
IntroductiontoProbability
PMFandCDFbothtermsbelongstoprobabilityandstatistics.Nowthequestionthatshould
ariseinyourmind,isthatwhyarewestudyingprobability.Itisbecausethesetwoconcepts
ofPMFandCDFaregoingtobeusedinthenexttutorialofHistogramequalization.Soifyou
dontknowhowtocalculatePMFandCDF,youcannotapplyhistogramequalizationonyour
image
.
WhatisPMF?
PMF stands for probability mass function. As it name suggest , it gives the probability of
eachnumberinthedatasetoryoucansaythatitbasicallygivesthecountorfrequencyof
eachelement.
HowPMFiscalculated:
We will calculate PMF from two different ways. First from a matrix , because in the next
tutorial,wehavetocalculatethePMFfromamatrix,andanimageisnothingmorethena
twodimensionalmatrix.
ThenwewilltakeanotherexampleinwhichwewillcalculatePMFfromthehistogram.
Considerthismatrix.
1
109/179
10/26/2016
DIPQuickGuide
NowifweweretocalculatethePMFofthismatrix,herehowwearegoingtodoit.
Atfirst,wewilltakethefirstvalueinthematrix,andthenwewillcount,howmuchtime
this value appears in the whole matrix. After count they can either be represented in a
histogram,orinatablelikethisbelow.
PMF
0
2/25
4/25
3/25
3/25
2/25
4/25
3/25
4/25
Notethatthesumofthecountmustbeequaltototalnumberofvalues.
CalculatingPMFfromhistogram
Theabovehistogramshowsfrequencyofgraylevelvaluesforan8bitsperpixelimage.
Now if we have to calculate its PMF , we will simple look at the count of each bar from
verticalaxisandthendivideitbytotalcount.
SothePMFoftheabovehistogramisthis.
110/179
10/26/2016
DIPQuickGuide
Another important thing to note in the above histogram is that it is not monotonically
increasing.Soinordertoincreaseitmonotonically,wewillcalculateitsCDF.
WhatisCDF?
CDF stands for cumulative distributive function. It is a function that calculates the
cumulativesumofallthevaluesthatarecalculatedbyPMF.Itbasicallysumstheprevious
one.
Howitiscalculated?
WewillcalculateCDFusingahistogram.Herehowitisdone.Considerthehistogramshown
abovewhichshowsPMF.
Sincethishistogramisnotincreasingmonotonically,sowillmakeitgrowmonotonically.
Wewillsimplykeepthefirstvalueasitis,andtheninthe2ndvalue,wewilladdthefirst
oneandsoon.
HereistheCDFoftheabovePMFfunction.
111/179
10/26/2016
DIPQuickGuide
Nowasyoucanseefromthegraphabove,thatthefirstvalueofPMFremainasitis.The
secondvalueofPMFisaddedinthefirstvalueandplacedover128.ThethirdvalueofPMF
isaddedinthesecondvalueofCDF,thatgives110/110whichisequalto1.
And also now , the function is growing monotonically which is necessary condition for
histogramequalization.
PMFandCDFusageinhistogramequalization
Histogramequalization.
Histogramequalizationisdiscussedinthenexttutorialbutabriefintroductionofhistogram
equalizationisgivenbelow.
Histogramequalizationisusedforenhancingthecontrastoftheimages.
PMF and CDF are both use in histogram equalization as it is described in the beginning of
thistutorial.Inthehistogramequalization,thefirstandthesecondsteparePMFandCDF.
Since in histogram equalization , we have to equalize all the pixel values of an image. So
PMFhelpsuscalculatingtheprobabilityofeachpixelvalueinanimage.AndCDFgivesus
thecumulativesumofthesevalues.Furtheron,thisCDFismultipliedbylevels,tofindthe
newpixelintensities,whicharemappedintooldvalues,andyourhistogramisequalized.
HistogramEqualization
We have already seen that contrast can be increased using histogram stretching. In this
tutorialwewillseethathowhistogramequalizationcanbeusedtoenhancecontrast.
Before performing histogram equalization, you must know two important concepts used in
equalizinghistograms.ThesetwoconceptsareknownasPMFandCDF.
TheyarediscussedinourtutorialofPMFandCDF.Pleasevisittheminordertosuccessfully
grasptheconceptofhistogramequalization.
HistogramEqualization:
Histogram equalization is used to enhance contrast. It is not necessary that contrast will
always be increase in this. There may be some cases were histogram equalization can be
worse.Inthatcasesthecontrastisdecreased.
Letsstarthistogramequalizationbytakingthisimagebelowasasimpleimage.
Image
112/179
10/26/2016
DIPQuickGuide
Histogramofthisimage:
Thehistogramofthisimagehasbeenshownbelow.
Nowwewillperformhistogramequalizationtoit.
PMF:
FirstwehavetocalculatethePMF(probabilitymassfunction)ofallthepixelsinthisimage.
IfyoudonotknowhowtocalculatePMF,pleasevisitourtutorialofPMFcalculation.
CDF:
Our next step involves calculation of CDF (cumulative distributive function). Again if you
donotknowhowtocalculateCDF,pleasevisitourtutorialofCDFcalculation.
CalculateCDFaccordingtograylevels
Letsforinstanceconsiderthis,thattheCDFcalculatedinthesecondsteplookslikethis.
113/179
10/26/2016
DIPQuickGuide
GrayLevelValue
CDF
0.11
0.22
0.55
0.66
0.77
0.88
0.99
TheninthisstepyouwillmultiplytheCDFvaluewith(Graylevels(minus)1).
Consideringwehavean3bppimage.Thennumberoflevelswehaveare8.And1subtracts
8is7.SowemultiplyCDFby7.Herewhatwegotaftermultiplying.
GrayLevelValue
CDF
CDF*(Levels1)
0.11
0.22
0.55
0.66
0.77
0.88
0.99
Now we have is the last step , in which we have to map the new gray level values into
numberofpixels.
Letsassumeouroldgraylevelsvalueshasthesenumberofpixels.
GrayLevelValue
Frequency
114/179
10/26/2016
DIPQuickGuide
10
12
14
16
Nowifwemapournewvaluesto,thenthisiswhatwegot.
GrayLevelValue
NewGrayLevelValue
Frequency
10
12
14
16
Nowmapthesenewvaluesyouareontohistogram,andyouaredone.
Lets apply this technique to our original image. After applying we got the following image
anditsfollowinghistogram.
HistogramEqualizationImage
115/179
10/26/2016
DIPQuickGuide
CumulativeDistributivefunctionofthisimage
HistogramEqualizationhistogram
Comparingboththehistogramsandimages
116/179
10/26/2016
DIPQuickGuide
Conclusion
Asyoucanclearlyseefromtheimagesthatthenewimagecontrasthasbeenenhancedand
itshistogramhasalsobeenequalized.Thereisalsooneimportantthingtobenoteherethat
during histogram equalization the overall shape of the histogram changes, where as in
histogramstretchingtheoverallshapeofhistogramremainssame.
GrayLevelTransformation
WehavediscussedsomeofthebasictransformationsinourtutorialofBasictransformation.
Inthistutorialwewilllookatsomeofthebasicgrayleveltransformations.
Imageenhancement
Enhancinganimageprovidesbettercontrastandamoredetailedimageascomparetonon
enhancedimage.Imageenhancementhasveryapplications.Itisusedtoenhancemedical
images,imagescapturedinremotesensing,imagesfromsatellitee.t.c
117/179
10/26/2016
DIPQuickGuide
Thetransformationfunctionhasbeengivenbelow
s=T(r)
where r is the pixels of the input image and s is the pixels of the output image. T is a
transformationfunctionthatmapseachvalueofrtoeachvalueofs.Imageenhancement
canbedonethroughgrayleveltransformationswhicharediscussedbelow.
Grayleveltransformation
Therearethreebasicgrayleveltransformation.
Linear
Logarithmic
Powerlaw
Theoverallgraphofthesetransitionshasbeenshownbelow.
Lineartransformation
Firstwewilllookatthelineartransformation.Lineartransformationincludessimpleidentity
and negative transformation. Identity transformation has been discussed in our tutorial of
imagetransformation,butabriefdescriptionofthistransformationhasbeengivenhere.
Identity transition is shown by a straight line. In this transition, each value of the input
image is directly mapped to each other value of output image. That results in the same
input image and output image. And hence is called identity transformation. It has been
shownbelow
118/179
10/26/2016
DIPQuickGuide
Negativetransformation
The second linear transformation is negative transformation, which is invert of identity
transformation.Innegativetransformation,eachvalueoftheinputimageissubtractedfrom
theL1andmappedontotheoutputimage.
Theresultissomewhatlikethis.
InputImage
OutputImage
Inthiscasethefollowingtransitionhasbeendone.
s=(L1)r
119/179
10/26/2016
DIPQuickGuide
sincetheinputimageofEinsteinisan8bppimage,sothenumberoflevelsinthisimage
are256.Putting256intheequation,wegetthis
s=255r
Soeachvalueissubtractedby255andtheresultimagehasbeenshownabove.Sowhat
happensisthat,thelighterpixelsbecomedarkandthedarkerpicturebecomeslight.Andit
resultsinimagenegative.
Ithasbeenshowninthegraphbelow.
Logarithmictransformations:
Logarithmictransformationfurthercontainstwotypeoftransformation.Logtransformation
andinverselogtransformation.
Logtransformation
Thelogtransformationscanbedefinedbythisformula
s=clog(r+1).
Wheresandrarethepixelvaluesoftheoutputandtheinputimageandcisaconstant.
Thevalue1isaddedtoeachofthepixelvalueoftheinputimagebecauseifthereisapixel
intensity of 0 in the image, then log (0) is equal to infinity. So 1 is added , to make the
minimumvalueatleast1.
During log transformation , the dark pixels in an image are expanded as compare to the
higher pixel values. The higher pixel values are kind of compressed in log transformation.
Thisresultinfollowingimageenhancement.
Thevalueofcinthelogtransformadjustthekindofenhancementyouarelookingfor.
120/179
10/26/2016
DIPQuickGuide
InputImage
LogTranformImage
Theinverselogtransformisoppositetologtransform.
PowerLawtransformations
There are further two transformation is power law transformations, that include nth power
andnthroottransformation.Thesetransformationscanbegivenbytheexpression:
s=cr^
Thissymboliscalledgamma,duetowhichthistransformationisalsoknownasgamma
transformation.
Variationinthevalueofvariestheenhancementoftheimages.Differentdisplaydevices/
monitorshavetheirowngammacorrection,thatswhytheydisplaytheirimageatdifferent
intensity.
This type of transformation is used for enhancing images for different type of display
devices. The gamma of different display devices is different. For example Gamma of CRT
liesinbetweenof1.8to2.5,thatmeanstheimagedisplayedonCRTisdark.
Correctinggamma.
s=cr^
s=cr^(1/2.5)
121/179
10/26/2016
DIPQuickGuide
Thesameimagebutwithdifferentgammavalueshasbeenshownhere.
Forexample:
Gamma=10
Gamma=8
Gamma=6
ConceptofConvolution
This tutorial is about one of the very important concept of signals and system. We will
completelydiscussconvolution.Whatisit?Whyisit?Whatcanweachievewithit?
Wewillstartdiscussingconvolutionfromthebasicsofimageprocessing.
Whatisimageprocessing.
Aswehavediscussedintheintroductiontoimageprocessingtutorialsandinthesignaland
systemthatimageprocessingismoreorlessthestudyofsignalsandsystemsbecausean
122/179
10/26/2016
DIPQuickGuide
imageisnothingbutatwodimensionalsignal.
Alsowehavediscussed,thatinimageprocessing,wearedevelopingasystemwhoseinput
isanimageandoutputwouldbeanimage.Thisispictoriallyrepresentedas.
TheboxisthatisshownintheabovefigurelabeledasDigitalImageProcessingsystem
couldbethoughtofasablackbox
Itcanbebetterrepresentedas:
Wherehavewereacheduntilnow
Tillnowwehavediscussedtwoimportantmethodstomanipulateimages.Orinotherwords
wecansaythat,ourblackboxworksintwodifferentwaystillnow.
Thetwodifferentwaysofmanipulatingimageswere
Graphs(Histograms)
This method is known as histogram processing. We have discussed it in detail in previous

tutorialsforincreasecontrast,imageenhancement,brightnesse.t.c
123/179
10/26/2016
DIPQuickGuide
Transformationfunctions
This method is known as transformations , in which we discussed different type of

transformationsandsomegrayleveltransformations
Anotherwayofdealingimages
Herewearegoingtodiscussanothermethodofdealingwithimages.Thisothermethodis
known as convolution. Usually the black box(system) used for image processing is an LTI
systemorlineartimeinvariantsystem.Bylinearwemeanthatsuchasystemwhereoutput
is always linear , neither log nor exponent or any other. And by time invariant we means
thatasystemwhichremainssameduringtime.
Sonowwearegoingtousethisthirdmethod.Itcanberepresentedas.
Itcanbemathematicallyrepresentedastwoways
g(x,y)=h(x,y)*f(x,y)
Itcanbeexplainedasthemaskconvolvedwithanimage.
Or
g(x,y)=f(x,y)*h(x,y)
Itcanbeexplainedasimageconvolvedwithmask.
Therearetwowaystorepresentthisbecausetheconvolutionoperator(*)iscommutative.
Theh(x,y)isthemaskorfilter.
124/179
10/26/2016
DIPQuickGuide
Whatismask?
Mask is also a signal. It can be represented by a two dimensional matrix. The mask is
usually of the order of 1x1, 3x3, 5x5 , 7x7 . A mask should always be in odd number ,
becauseotherwiseyoucannotfindthemidofthemask.Whydoweneedtofindthemidof
themask.Theanswerliesbelow,intopicof,howtoperformconvolution?
Howtoperformconvolution?
Inordertoperformconvolutiononanimage,followingstepsshouldbetaken.
Flipthemask(horizontallyandvertically)onlyonce
Slidethemaskontotheimage.
Multiplythecorrespondingelementsandthenaddthem
Repeatthisprocedureuntilallvaluesoftheimagehasbeencalculated.
Exampleofconvolution
Letsperformsomeconvolution.Step1istoflipthemask.
Mask:
Letstakeourmasktobethis.
1
Flippingthemaskhorizontally
Flippingthemaskvertically
125/179
10/26/2016
DIPQuickGuide
Image:
Letsconsideranimagetobelikethis
2
10
12
14
16
18
Convolution
Convolvingmaskoverimage.Itisdoneinthisway.Placethecenterofthemaskateach
elementofanimage.Multiplythecorrespondingelementsandthenaddthem,andpaste
theresultontotheelementoftheimageonwhichyouplacethecenterofmask.
Theboxinredcoloristhemask,andthevaluesintheorangearethevaluesofthemask.
Theblackcolorboxandvaluesbelongtotheimage.Nowforthefirstpixeloftheimage,the
valuewillbecalculatedas
Firstpixel=(5*2)+(4*4)+(2*8)+(1*10)
=10+16+16+10
=52
Place52intheoriginalimageatthefirstindexandrepeatthisprocedureforeachpixelof
theimage.
WhyConvolution
Convolutioncanachievesomething,thattheprevioustwomethodsofmanipulatingimages
cantachieve.Thoseincludetheblurring,sharpening,edgedetection,noisereductione.t.c
ConceptofMask
Whatisamask.
126/179
10/26/2016
DIPQuickGuide
A mask is a filter. Concept of masking is also known as spatial filtering. Masking is also
knownasfiltering.Inthisconceptwejustdealwiththefilteringoperationthatisperformed
directlyontheimage.
Asamplemaskhasbeenshownbelow
1
Whatisfiltering.
Theprocessoffilteringisalsoknownasconvolvingamaskwithanimage.Asthisprocessis
sameofconvolutionsofiltermasksarealsoknownasconvolutionmasks.
Howitisdone.
The general process of filtering and applying masks is consists of moving the filter mask
frompointtopointinanimage.Ateachpoint(x,y)oftheoriginalimage,theresponseofa
filteriscalculatedbyapredefinedrelationship.Allthefiltersvaluesarepredefinedandare
astandard.
Typesoffilters
Generallytherearetwotypesoffilters.Oneiscalledaslinearfiltersorsmoothingfiltersand
othersarecalledasfrequencydomainfilters.
Whyfiltersareused?
Filters are applied on image for multiple purposes. The two most common uses are as
following:
FiltersareusedforBlurringandnoisereduction
Filtersareusedoredgedetectionandsharpness
Blurringandnoisereduction:
Filtersaremostcommonlyusedforblurringandfornoisereduction.Blurringisusedinpre
processing steps, such as removal of small details from an image prior to large object
extraction.
Masksforblurring.
Thecommonmasksforblurringare.
Boxfilter
127/179
10/26/2016
DIPQuickGuide
Weightedaveragefilter
In the process of blurring we reduce the edge content in an image and try to make the
transitionsbetweendifferentpixelintensitiesassmoothaspossible.
Noisereductionisalsopossiblewiththehelpofblurring.
EdgeDetectionandsharpness:
Masksorfilterscanalsobeusedforedgedetectioninanimageandtoincreasesharpness
ofanimage.
Whatareedges.
We can also say that sudden changes of discontinuities in an image are called as edges.
Significanttransitionsinanimagearecalledasedges.Apicturewithedgesisshownbelow.
Originalpicture.
Samepicturewithedges
ConceptofBlurring
A brief introduction of blurring has been discussed in our previous tutorial of concept of
masks,butweareformallygoingtodiscussithere.
Blurring
128/179
10/26/2016
DIPQuickGuide
Inblurring,wesimplebluranimage.Animagelooksmoresharpormoredetailedifweare
abletoperceivealltheobjectsandtheirshapescorrectlyinit.Forexample.Animagewith
aface,looksclearwhenweareabletoidentifyeyes,ears,nose,lips,foreheade.t.cvery
clear.Thisshapeofanobjectisduetoitsedges.Soinblurring,wesimplereducetheedge
contentandmakesthetransitionformonecolortotheotherverysmooth.
Blurringvszooming.
Youmighthaveseenablurredimagewhenyouzoomanimage.Whenyouzoomanimage
using pixel replication , and zooming factor is increased, you saw a blurred image. This
imagealsohaslessdetails,butitisnottrueblurring.
Becauseinzooming,youaddnewpixelstoanimage,thatincreasetheoverallnumberof
pixels in an image , whereas in blurring , the number of pixels of a normal image and a
blurredimageremainsthesame.
Commonexampleofablurredimage.
Typesoffilters.
Blurring can be achieved by many ways. The common type of filters that are used to
performblurringare.
Meanfilter
Weightedaveragefilter
Gaussianfilter
Out of these three , we are going to discuss the first two here and Gaussian will be
discussedlateronintheupcomingtutorials.
Meanfilter.
Mean filter is also known as Box filter and average filter. A mean filter has the following
properties.
Itmustbeoddordered
129/179
10/26/2016
DIPQuickGuide
Thesumofalltheelementsshouldbe1
Alltheelementsshouldbesame
Ifwefollowthisrule,thenforamaskof3x3.Wegetthefollowingresult.
1/9
1/9
1/9
1/9
1/9
1/9
1/9
1/9
1/9
Since it is a 3x3 mask, that means it has 9 cells. The condition that all the element sum
shouldbeequalto1canbeachievedbydividingeachvalueby9.As
1/9+1/9+1/9+1/9+1/9+1/9+1/9+1/9+1/9=9/9=1
Theresultofamaskof3x3onanimageisshownbelow.
OriginalImage:
BlurredImage
May be the results are not much clear. Lets increase the blurring. The blurring can be
increasedbyincreasingthesizeofthemask.Themoreisthesizeofthemask,themoreis
the blurring. Because with greater mask , greater number of pixels are catered and one
smoothtransitionisdefined.
130/179
10/26/2016
DIPQuickGuide
OriginalImage:
BlurredImage:
Samewayifweincreasethemask,theblurringwouldbemoreandtheresultsareshown
below.
OriginalImage:
BlurredImage:
131/179
10/26/2016
DIPQuickGuide
OriginalImage:
BlurredImage:
OriginalImage:
BlurredImage:
Weightedaveragefilter.
132/179
10/26/2016
DIPQuickGuide
In weighted average filter, we gave more weight to the center value. Due to which the
contributionofcenterbecomesmorethentherestofthevalues.Duetoweightedaverage
filtering,wecanactuallycontroltheblurring.
Propertiesoftheweightedaveragefilterare.
Itmustbeoddordered
Thesumofalltheelementsshouldbe1
Theweightofcenterelementshouldbemorethenalloftheotherelements
Filter1
1
Thetwopropertiesaresatisfiedwhichare(1and3).Buttheproperty2isnotsatisfied.So
inordertosatisfythatwewillsimpledividethewholefilterby10,ormultiplyitwith1/10.
Filter2
1
10
Dividingfactor=18.
ConceptofEdgeDetection
Wehavediscussedbrieflyaboutedgedetectioninourtutorialofintroductiontomasks.We
willformallydiscussedgedetectionhere.
Whatareedges.
We can also say that sudden changes of discontinuities in an image are called as edges.
Significanttransitionsinanimagearecalledasedges.
Typesofedges.
Geenerallyedgesareofthreetypes:
133/179
10/26/2016
DIPQuickGuide
Horizontaledges
VerticalEdges
DiagonalEdges
Whydetectedges.
Most of the shape information of an image is enclosed in edges. So first we detect these
edgesinanimageandbyusingthesefiltersandthenbyenhancingthoseareasofimage
whichcontainsedges,sharpnessoftheimagewillincreaseandimagewillbecomeclearer.
Here are some of the masks for edge detection that we will discuss in the upcoming
tutorials.
PrewittOperator
SobelOperator
RobinsonCompassMasks
KrischCompassMasks
LaplacianOperator.
AbovementionedallthefiltersareLinearfiltersorsmoothingfilters.
PrewittOperator
Prewittoperatorisusedfordetectingedgeshorizontallyandvertically.
SobelOperator
ThesobeloperatorisverysimilartoPrewittoperator.Itisalsoaderivatemaskandisused
foredgedetection.Italsocalculatesedgesinbothhorizontalandverticaldirection.
RobinsonCompassMasks
Thisoperatorisalsoknownasdirectionmask.Inthisoperatorwetakeonemaskandrotate
itinallthe8compassmajordirectionstocalculateedgesofeachdirection.
KirschCompassMasks
KirschCompassMaskisalsoaderivativemaskwhichisusedforfindingedges.Kirschmask
isalsousedforcalculatingedgesinallthedirections.
LaplacianOperator.
Laplacian Operator is also a derivative operator which is used to find edges in an image.
Laplacianisasecondorderderivativemask.Itcanbefurtherdividedintopositivelaplacian
134/179
10/26/2016
DIPQuickGuide
andnegativelaplacian.
Allthesemasksfindedges.Somefindhorizontallyandvertically,somefindinonedirection
onlyandsomefindinallthedirections.Thenextconceptthatcomesafterthisissharpening
whichcanbedoneoncetheedgesareextractedfromtheimage
Sharpening:
Sharpening is opposite to the blurring. In blurring, we reduce the edge content and in
sharpneng , we increase the edge content. So in order to increase the edge content in an
image,wehavetofindedgesfirst.
Edgescanbefindbyoneoftheanymethoddescribedabovebyusinganyoperator.After
findingedges,wewilladdthoseedgesonanimageandthustheimagewouldhavemore
edges,anditwouldlooksharpen.
Thisisonewayofsharpeninganimage.
Thesharpenimageisshownbelow.
OriginalImage
SharpenImage
PrewittOperator
Prewittoperatorisusedforedgedetectioninanimage.Itdetectstwotypesofedges:
135/179
10/26/2016
DIPQuickGuide
Horizontaledges
VerticalEdges
Edges are calculated by using difference between corresponding pixel intensities of an
image.Allthemasksthatareusedforedgedetectionarealsoknownasderivativemasks.
Becauseaswehavestatedmanytimesbeforeinthisseriesoftutorialsthatimageisalsoa
signalsochangesinasignalcanonlybecalculatedusingdifferentiation.Sothatswhythese
operatorsarealsocalledasderivativeoperatorsorderivativemasks.
Allthederivativemasksshouldhavethefollowingproperties:
Oppositesignshouldbepresentinthemask.
Sumofmaskshouldbeequaltozero.
Moreweightmeansmoreedgedetection.
Prewittoperatorprovidesustwomasksonefordetectingedgesinhorizontaldirectionand
anotherfordetectingedgesinanverticaldirection.
Verticaldirection:
1
Abovemaskwillfindtheedgesinverticaldirectionanditisbecausethezeroscolumninthe
verticaldirection.Whenyouwillconvolvethismaskonanimage,itwillgiveyouthevertical
edgesinanimage.
Howitworks:
Whenweapplythismaskontheimageitprominentverticaledges.Itsimplyworkslikeas
firstorderderivateandcalculatesthedifferenceofpixelintensitiesinaedgeregion.Asthe
centercolumnisofzerosoitdoesnotincludetheoriginalvaluesofanimagebutratherit
calculates the difference of right and left pixel values around that edge. This increase the
edgeintensityanditbecomeenhancedcomparativelytotheoriginalimage.
HorizontalDirection:
1
136/179
10/26/2016
DIPQuickGuide
Abovemaskwillfindedgesinhorizontaldirectionanditisbecausethatzeroscolumnisin
horizontal direction. When you will convolve this mask onto an image it would prominent
horizontaledgesintheimage.
Howitworks:
Thismaskwillprominentthehorizontaledgesinanimage.Italsoworksontheprincipleof
above mask and calculates difference among the pixel intensities of a particular edge. As
thecenterrowofmaskisconsistofzerossoitdoesnotincludetheoriginalvaluesofedge
intheimagebutratheritcalculatethedifferenceofaboveandbelowpixelintensitiesofthe
particularedge.Thusincreasingthesuddenchangeofintensitiesandmakingtheedgemore
visible. Both the above masks follow the principle of derivate mask. Both masks have
opposite sign in them and both masks sum equals to zero. The third condition will not be
applicableinthisoperatorasboththeabovemasksarestandardizeandwecantchangethe
valueinthem.
Nowitstimetoseethesemasksinaction:
SampleImage:
Followingisasamplepictureonwhichwewillapplyabovetwomasksoneattime.
AfterapplyingVerticalMask:
Afterapplyingverticalmaskontheabovesampleimage,followingimagewillbeobtained.
This image contains vertical edges. You can judge it more correctly by comparing with
horizontaledgespicture.
137/179
10/26/2016
DIPQuickGuide
AfterapplyingHorizontalMask:
Afterapplyinghorizontalmaskontheabovesampleimage,followingimagewillbeobtained.
Comparison:
As you can see that in the first picture on which we apply vertical mask, all the vertical
edges are more visible than the original image. Similarly in the second picture we have
appliedthehorizontalmaskandinresultallthehorizontaledgesarevisible.Sointhisway
youcanseethatwecandetectbothhorizontalandverticaledgesfromanimage.
SobelOperator
ThesobeloperatorisverysimilartoPrewittoperator.Itisalsoaderivatemaskandisused
foredgedetection.LikePrewittoperatorsobeloperatorisalsousedtodetecttwokindsof
edgesinanimage:
138/179
10/26/2016
DIPQuickGuide
Verticaldirection
Horizontaldirection
DifferencewithPrewittOperator:
The major difference is that in sobel operator the coefficients of masks are not fixed and
theycanbeadjustedaccordingtoourrequirementunlesstheydonotviolateanyproperty
ofderivativemasks.
FollowingistheverticalMaskofSobelOperator:
1
This mask works exactly same as the Prewitt operator vertical mask. There is only one
differencethatisithas2and2valuesincenteroffirstandthirdcolumn.Whenapplied
onanimagethismaskwillhighlighttheverticaledges.
Howitworks:
Whenweapplythismaskontheimageitprominentverticaledges.Itsimplyworkslikeas
firstorderderivateandcalculatesthedifferenceofpixelintensitiesinaedgeregion.
As the center column is of zero so it does not include the original values of an image but
rather it calculates the difference of right and left pixel values around that edge. Also the
centervaluesofboththefirstandthirdcolumnis2and2respectively.
This give more weight age to the pixel values around the edge region. This increase the
edgeintensityanditbecomeenhancedcomparativelytotheoriginalimage.
FollowingisthehorizontalMaskofSobelOperator:
1
Abovemaskwillfindedgesinhorizontaldirectionanditisbecausethatzeroscolumnisin
horizontal direction. When you will convolve this mask onto an image it would prominent
horizontaledgesintheimage.Theonlydifferencebetweenitisthatithave2and2asa
centerelementoffirstandthirdrow.
139/179
10/26/2016
DIPQuickGuide
Howitworks:
Thismaskwillprominentthehorizontaledgesinanimage.Italsoworksontheprincipleof
above mask and calculates difference among the pixel intensities of a particular edge. As
thecenterrowofmaskisconsistofzerossoitdoesnotincludetheoriginalvaluesofedge
intheimagebutratheritcalculatethedifferenceofaboveandbelowpixelintensitiesofthe
particularedge.Thusincreasingthesuddenchangeofintensitiesandmakingtheedgemore
visible.
Nowitstimetoseethesemasksinaction:
SampleImage:
Followingisasamplepictureonwhichwewillapplyabovetwomasksoneattime.
AfterapplyingVerticalMask:
Afterapplyingverticalmaskontheabovesampleimage,followingimagewillbeobtained.
140/179
10/26/2016
DIPQuickGuide
AfterapplyingHorizontalMask:
Afterapplyinghorizontalmaskontheabovesampleimage,followingimagewillbeobtained
Comparison:
As you can see that in the first picture on which we apply vertical mask, all the vertical
edges are more visible than the original image. Similarly in the second picture we have
appliedthehorizontalmaskandinresultallthehorizontaledgesarevisible.
Sointhiswayyoucanseethatwecandetectbothhorizontalandverticaledgesfroman
image.AlsoifyoucomparetheresultofsobeloperatorwithPrewittoperator,youwillfind
that sobel operator finds more edges or make edges more visible as compared to Prewitt
Operator.
This is because in sobel operator we have allotted more weight to the pixel intensities
aroundtheedges.
Applyingmoreweighttomask
Nowwecanalsoseethatifweapplymoreweighttothemask,themoreedgesitwillget
for us. Also as mentioned in the start of the tutorial that there is no fixed coefficients in
sobeloperator,sohereisanotherweightedoperator
1
141/179
10/26/2016
DIPQuickGuide
IfyoucancomparetheresultofthismaskwithofthePrewittverticalmask,itisclearthat
this mask will give out more edges as compared to Prewitt one just because we have
allottedmoreweightinthemask.
RobinsonCompassMask
Robinson compass masks are another type of derrivate mask which is used for edge
detection.Thisoperatorisalsoknownasdirectionmask.Inthisoperatorwetakeonemask
androtateitinallthe8compassmajordirectionsthatarefollowing:
North
NorthWest
West
SouthWest
South
SouthEast
East
NorthEast
Thereisnofixedmask.Youcantakeanymaskandyouhavetorotateittofindedgesinall
theabovementioneddirections.Allthemasksarerotatedonthebasesofdirectionofzero
columns.
For example lets see the following mask which is in North Direction and then rotate it to
makeallthedirectionmasks.
NorthDirectionMask
1
NorthWestDirectionMask
0
142/179
10/26/2016
DIPQuickGuide
WestDirectionMask
1
SouthWestDirectionMask
2
SouthDirectionMask
1
SouthEastDirectionMask
0
EastDirectionMask
1
NorthEastDirectionMask
2
143/179
10/26/2016
DIPQuickGuide
Asyoucanseethatallthedirectionsarecoveredonthebasisofzerosdirection.Eachmask
willgiveyoutheedgesonitsdirection.Nowletsseetheresultoftheentireabovemasks.
Suppose we have a sample picture from which we have to find all the edges. Here is our
samplepicture:
SamplePicture:
Nowwewillapplyalltheabovefiltersonthisimageandwegetthefollowingresult.
NorthDirectionEdges
NorthWestDirectionEdges
WestDirectionEdges
144/179
10/26/2016
DIPQuickGuide
SouthWestDirectionEdges
SouthDirectionEdges
SouthEastDirectionEdges
EastDirectionEdges
145/179
10/26/2016
DIPQuickGuide
NorthEastDirectionEdges
Asyoucanseethatbyapplyingalltheabovemasksyouwillgetedgesinallthedirection.
Result is also depends on the image. Suppose there is an image, which do not have any
NorthEastdirectionedgessothenthatmaskwillbeineffective.
KrischCompassMask
KirschCompassMaskisalsoaderivativemaskwhichisusedforfindingedges.Thisisalso
like Robinson compass find edges in all the eight directions of a compass. The only
differencebetweenRobinsonandkirschcompassmasksisthatinKirschwehaveastandard
maskbutinKirschwechangethemaskaccordingtoourownrequirements.
WiththehelpofKirschCompassMaskswecanfindedgesinthefollowingeightdirections.
North
NorthWest
West
SouthWest
South
SouthEast
East
146/179
10/26/2016
DIPQuickGuide
NorthEast
We take a standard mask which follows all the properties of a derivative mask and then
rotateittofindtheedges.
For example lets see the following mask which is in North Direction and then rotate it to
makeallthedirectionmasks.
NorthDirectionMask
3
NorthWestDirectionMask
3
WestDirectionMask
SouthWestDirectionMask
5
SouthDirectionMask
5
147/179
10/26/2016
DIPQuickGuide
SouthEastDirectionMask
3
EastDirectionMask
NorthEastDirectionMask
Asyoucanseethatallthedirectionsarecoveredandeachmaskwillgiveyoutheedgesof
its own direction. Now to help you better understand the concept of these masks we will
applyitonarealimage.Supposewehaveasamplepicturefromwhichwehavetofindall
theedges.Hereisoursamplepicture:
SamplePicture
Nowwewillapplyalltheabovefiltersonthisimageandwegetthefollowingresult.
NorthDirectionEdges
148/179
10/26/2016
DIPQuickGuide
NorthWestDirectionEdges
WestDirectionEdges
SouthWestDirectionEdges
SouthDirectionEdges
149/179
10/26/2016
DIPQuickGuide
SouthEastDirectionEdges
EastDirectionEdges
NorthEastDirectionEdges
Asyoucanseethatbyapplyingalltheabovemasksyouwillgetedgesinallthedirection.
Result is also depends on the image. Suppose there is an image, which do not have any
NorthEastdirectionedgessothenthatmaskwillbeineffective.
LaplacianOperator
150/179
LaplacianOperator
10/26/2016
DIPQuickGuide
LaplacianOperatorisalsoaderivativeoperatorwhichisusedtofindedgesinanimage.The
major difference between Laplacian and other operators like Prewitt, Sobel, Robinson and
Kirsch is that these all are first order derivative masks but Laplacian is a second order
derivativemask.InthismaskwehavetwofurtherclassificationsoneisPositiveLaplacian
OperatorandotherisNegativeLaplacianOperator.
Another difference between Laplacian and other operators is that unlike other operators
Laplaciandidnttakeoutedgesinanyparticulardirectionbutittakeoutedgesinfollowing
classification.
InwardEdges
OutwardEdges
LetsseethathowLaplacianoperatorworks.
PositiveLaplacianOperator:
InPositiveLaplacianwehavestandardmaskinwhichcenterelementofthemaskshouldbe
negativeandcornerelementsofmaskshouldbezero.
0
PositiveLaplacianOperatorisusetotakeoutoutwardedgesinanimage.
NegativeLaplacianOperator:
In negative Laplacian operator we also have a standard mask, in which center element
shouldbepositive.Alltheelementsinthecornershouldbezeroandrestofalltheelements
inthemaskshouldbe1.
0
NegativeLaplacianoperatorisusetotakeoutinwardedgesinanimage
Howitworks:
151/179
10/26/2016
DIPQuickGuide
Laplacian is a derivative operator its uses highlight gray level discontinuities in an image
and try to deemphasize regions with slowly varying gray levels. This operation in result
produces such images which have grayish edge lines and other discontinuities on a dark
background.Thisproducesinwardandoutwardedgesinanimage
Theimportantthingishowtoapplythesefiltersontoimage.Rememberwecantapplyboth
thepositiveandnegativeLaplacianoperatoronthesameimage.wehavetoapplyjustone
butthethingtorememberisthatifweapplypositiveLaplacianoperatorontheimagethen
we subtract the resultant image from the original image to get the sharpened image.
SimilarlyifweapplynegativeLaplacianoperatorthenwehavetoaddtheresultantimage
ontooriginalimagetogetthesharpenedimage.
Letsapplythesefiltersontoanimageandseehowitwillgetusinwardandoutwardedges
fromanimage.Supposewehaveafollowingsampleimage.
SampleImage
AfterapplyingPositiveLaplacianOperator:
AfterapplyingpositiveLaplacianoperatorwewillgetthefollowingimage.
152/179
10/26/2016
DIPQuickGuide
AfterapplyingNegativeLaplacianOperator:
AfterapplyingnegativeLaplacianoperatorwewillgetthefollowingimage.
IntroductiontoFrequencydomain
We have deal with images in many domains. Now we are processing signals (images) in
frequencydomain.SincethisFourierseriesandfrequencydomainispurelymathematics,
sowewilltrytominimizethatmathspartandfocusmoreonitsuseinDIP.
Frequencydomainanalysis
Tillnow,allthedomainsinwhichwehaveanalyzedasignal,weanalyzeitwithrespectto
time. But in frequency domain we dont analyze signal with respect to time , but with
respectoffrequency.
153/179
10/26/2016
DIPQuickGuide
Differencebetweenspatialdomainandfrequencydomain.
Inspatialdomain,wedealwithimagesasitis.Thevalueofthepixelsoftheimagechange
with respect to scene. Whereas in frequency domain , we deal with the rate at which the
pixelvaluesarechanginginspatialdomain.
Forsimplicity,Letsputitthisway.
Spatialdomain
In simple spatial domain , we directly deal with the image matrix. Whereas in frequency
domain,wedealanimagelikethis.
FrequencyDomain
We first transform the image to its frequency distribution. Then our black box system
performwhateverprocessingithastoperformed,andtheoutputoftheblackboxinthis
caseisnotanimage,butatransformation.Afterperforminginversetransformation,itis
convertedintoanimagewhichisthenviewedinspatialdomain.
Itcanbepictoriallyviewedas
Herewehaveusedthewordtransformation.Whatdoesitactuallymean?
Transformation.
A signal can be converted from time domain into frequency domain using mathematical
operatorscalledtransforms.Therearemanykindoftransformationthatdoesthis.Someof
154/179
10/26/2016
DIPQuickGuide
themaregivenbelow.
FourierSeries
Fouriertransformation
Laplacetransform
Ztransform
Outofallthese,wewillthoroughlydiscussFourierseriesandFouriertransformationinour
nexttutorial.
Frequencycomponents
Any image in spatial domain can be represented in a frequency domain. But what do this
frequenciesactuallymean.
Wewilldividefrequencycomponentsintotwomajorcomponents.
Highfrequencycomponents
Highfrequencycomponentscorrespondtoedgesinanimage.
Lowfrequencycomponents
Lowfrequencycomponentsinanimagecorrespondtosmoothregions.
FourierSeriesandTransform
In the last tutorial of Frequency domain analysis, we discussed that Fourier series and
Fouriertransformareusedtoconvertasignaltofrequencydomain.
Fourier
Fourier was a mathematician in 1822. He give Fourier series and Fourier transform to
convertasignalintofrequencydomain.
FourierSeries
Fourierseriessimplystatesthat,periodicsignalscanberepresentedintosumofsinesand
cosineswhenmultipliedwithacertainweight.Itfurtherstatesthatperiodicsignalscanbe
brokendownintofurthersignalswiththefollowingproperties.
Thesignalsaresinesandcosines
Thesignalsareharmonicsofeachother
Itcanbepictoriallyviewedas
155/179
10/26/2016
DIPQuickGuide
Intheabovesignal,thelastsignalisactuallythesumofalltheabovesignals.Thiswasthe
ideaoftheFourier.
Howitiscalculated.
Since as we have seen in the frequency domain , that in order to process an image in
frequencydomain,weneedtofirstconvertitusingintofrequencydomainandwehaveto
take inverse of the output to convert it back into spatial domain. Thats why both Fourier
series and Fourier transform has two formulas. One for conversion and one converting it
backtothespatialdomain.
Fourierseries
TheFourierseriescanbedenotedbythisformula.
Theinversecanbecalculatedbythisformula.
Fouriertransform
TheFouriertransformsimplystatesthatthatthenonperiodicsignalswhoseareaunderthe
curve is finite can also be represented into integrals of the sines and cosines after being
multipliedbyacertainweight.
The Fourier transform has many wide applications that include , image compression (e.g
JPEGcompression),filtreringandimageanalysis.
156/179
10/26/2016
DIPQuickGuide
DifferencebetweenFourierseriesandtransform
AlthoughbothFourierseriesandFouriertransformaregivenbyFourier,butthedifference
betweenthemisFourierseriesisappliedonperiodicsignalsandFouriertransformisapplied
fornonperiodicsignals
Whichoneisappliedonimages.
Now the question is that which one is applied on the images , the Fourier series or the
Fouriertransform.Well,theanswertothisquestionliesinthefactthatwhatimagesare.
Imagesarenonperiodic.Andsincetheimagesarenonperiodic,soFouriertransformis
usedtoconvertthemintofrequencydomain.
Discretefouriertransform.
Sincewearedealingwithimages,andinfactdigitalimages,sofordigitalimageswewillbe
workingondiscretefouriertransform
ConsidertheaboveFouriertermofasinusoid.Itincludethreethings.
SpatialFrequency
Magnitude
Phase
Thespatialfrequencydirectlyrelateswiththebrightnessoftheimage.Themagnitudeofthe
sinusoiddirectlyrelateswiththecontrast.Contrastisthedifferencebetweenmaximumand
minimumpixelintensity.Phasecontainsthecolorinformation.
Theformulafor2dimensionaldiscreteFouriertransformisgivenbelow.
The discrete Fourier transform is actually the sampled Fourier transform, so it contains
somesamplesthatdenotesanimage.Intheaboveformulaf(x,y)denotestheimage,and
F(u,v) denotes the discrete Fourier transform. The formula for 2 dimensional inverse
discreteFouriertransformisgivenbelow.
157/179
10/26/2016
DIPQuickGuide
TheinversediscreteFouriertransformconvertstheFouriertransformbacktotheimage
Considerthissignal.
Now we will see an image , whose we will calculate FFT magnitude spectrum and then
shiftedFFTmagnitudespectrumandthenwewilltakeLogofthatshiftedspectrum.
OriginalImage
TheFouriertransformmagnitudespectrum
TheShiftedFouriertransform
158/179
10/26/2016
DIPQuickGuide
TheShiftedMagnitudeSpectrum
ConvolutionTheorem
Inthelasttutorial,wediscussedabouttheimagesinfrequencydomain.Inthistutorial,we
are going to define a relationship between frequency domain and the images(spatial
domain).
Forexample:
Considerthisexample.
Thesameimageinthefrequencydomaincanberepresentedas.
159/179
10/26/2016
DIPQuickGuide
Now whats the relationship between image or spatial domain and frequency domain. This
relationshipcanbeexplainedbyatheoremwhichiscalledasConvolutiontheorem.
ConvolutionTheorem
Therelationshipbetweenthespatialdomainandthefrequencydomaincanbeestablished
byconvolutiontheorem.
Theconvolutiontheoremcanberepresentedas.
Itcanbestatedastheconvolutioninspatialdomainisequaltofilteringinfrequencydomain
andviceversa.
Thefilteringinfrequencydomaincanberepresentedasfollowing:
Thestepsinfilteringaregivenbelow.
At first step we have to do some pre processing an image in spatial domain,
meansincreaseitscontrastorbrightness
ThenwewilltakediscreteFouriertransformoftheimage
Then we will center the discrete Fourier transform , as we will bring the discrete
Fouriertransformincenterfromcorners
Thenwewillapplyfiltering,meanswewillmultiplytheFouriertransformbyafilter
function
ThenwewillagainshifttheDFTfromcentertothecorners
160/179
10/26/2016
DIPQuickGuide
Last step would be take to inverse discrete Fourier transform , to bring the result
backfromfrequencydomaintospatialdomain
Andthisstepofpostprocessingisoptional,justlikepreprocessing,inwhichwe
justincreasetheappearanceofimage.
Filters
Theconceptoffilterinfrequencydomainissameastheconceptofamaskinconvolution.
Afterconvertinganimagetofrequencydomain,somefiltersareappliedinfilteringprocess
to perform different kind of processing on an image. The processing include blurring an
image,sharpeninganimagee.t.c.
Thecommontypeoffiltersforthesepurposesare:
Idealhighpassfilter
Ideallowpassfilter
Gaussianhighpassfilter
Gaussianlowpassfilter
Inthenexttutorial,wewilldiscussaboutfilterindetail.
HighPassvsLowPassFilters
Inthelasttutorial,webrieflydiscussaboutfilters.Inthistutorialwewillthoroughlydiscuss
aboutthem.Beforediscussingaboutletstalkaboutmasksfirst.Theconceptofmaskhas
beendiscussedinourtutorialofconvolutionandmasks.
Blurringmasksvsderivativemasks.
Wearegoingtoperformacomparisonbetweenblurringmasksandderivativemasks.
Blurringmasks:
Ablurringmaskhasthefollowingproperties.
Allthevaluesinblurringmasksarepositive
Thesumofallthevaluesisequalto1
Theedgecontentisreducedbyusingablurringmask
Asthesizeofthemaskgrow,moresmoothingeffectwilltakeplace
161/179
10/26/2016
DIPQuickGuide
Derrivativemasks:
Aderivativemaskhasthefollowingproperties.
Aderivativemaskhavepositiveandaswellasnegativevalues
Thesumofallthevaluesinaderivativemaskisequaltozero
Theedgecontentisincreasedbyaderivativemask
Asthesizeofthemaskgrows,moreedgecontentisincreased
Relationshipbetweenblurringmaskandderivativemaskwithhighpass
filtersandlowpassfilters.
Therelationshipbetweenblurringmaskandderivativemaskwithahighpassfilterandlow
passfiltercanbedefinedsimplyas.
Blurringmasksarealsocalledaslowpassfilter
Derivativemasksarealsocalledashighpassfilter
HighpassfrequencycomponentsandLowpassfrequencycomponents
The high pass frequency components denotes edges whereas the low pass frequency
componentsdenotessmoothregions.
IdeallowpassandIdealHighpassfilters
Thisisthecommonexampleoflowpassfilter.
Whenoneisplacedinsideandthezeroisplacedoutside,wegotablurredimage.Nowas
we increase the size of 1, blurring would be increased and the edge content would be
reduced.
Thisisacommonexampleofhighpassfilter.
162/179
10/26/2016
DIPQuickGuide
When0isplacedinside,wegetedges,whichgivesusasketchedimage.Anideallowpass
filterinfrequencydomainisgivenbelow
Theideallowpassfiltercanbegraphicallyrepresentedas
Nowletsapplythisfiltertoanactualimageandletsseewhatwegot.
Sampleimage.
163/179
10/26/2016
DIPQuickGuide
Imageinfrequencydomain
Applyingfilteroverthisimage
ResultantImage
Withthesameway,anidealhighpassfiltercanbeappliedonanimage.Butobviouslythe
results would be different as , the low pass reduces the edged content and the high pass
increaseit.
164/179
10/26/2016
DIPQuickGuide
GaussianLowpassandGaussianHighpassfilter
Gaussian low pass and Gaussian high pass filter minimize the problem that occur in ideal
lowpassandhighpassfilter.
This problem is known as ringing effect. This is due to reason because at some points
transition between one color to the other cannot be defined precisely, due to which the
ringingeffectappearsatthatpoint.
Havealookatthisgraph.
Thisistherepresentationofideallowpassfilter.NowattheexactpointofDo,youcannot
tellthatthevaluewouldbe0or1.Duetowhichtheringingeffectappearsatthatpoint.
Soinordertoreducetheeffectthatappearsisideallowpassandidealhighpassfilter,the
followingGaussianlowpassfilterandGaussianhighpassfilterisintroduced.
GaussianLowpassfilter
The concept of filtering and low pass remains the same, but only the transition becomes
differentandbecomemoresmooth.
TheGaussianlowpassfiltercanberepresentedas
165/179
10/26/2016
DIPQuickGuide
Note the smooth curve transition, due to which at each point, the value of Do , can be
exactlydefined.
Gaussianhighpassfilter
Gaussian high pass filter has the same concept as ideal high pass filter , but again the
transitionismoresmoothascomparedtotheidealone.
IntroductiontoColorSpaces
Inthistutorial,wearegoingtotalkaboutcolorspaces.
Whatarecolorspaces?
Colorspacesaredifferenttypesofcolormodes,usedinimageprocessingandsignalsand
systemforvariouspurposes.Someofthecommoncolorspacesare:
RGB
CMYK
YUV
YIQ
YCbCr
HSV
RGB
166/179
10/26/2016
DIPQuickGuide
RGB is the most widely used color space , and we have already discussed it in the past
tutorials.RGBstandsforredgreenandblue.
WhatRGBmodelstates,thateachcolorimageisactuallyformedofthreedifferentimages.
Redimage,Blueimage,andblackimage.Anormalgrayscaleimagecanbedefinedbyonly
onematrix,butacolorimageisactuallycomposedofthreedifferentmatrices.
Onecolorimagematrix=redmatrix+bluematrix+greenmatrix
Thiscanbebestseeninthisexamplebelow.
ApplicationsofRGB
ThecommonapplicationsofRGBmodelare
Cathoderaytube(CRT)
Liquidcrystaldisplay(LCD)
167/179
10/26/2016
DIPQuickGuide
PlasmaDisplayorLEDdisplaysuchasatelevision
Acomputemonitororalargescalescreen
CMYK
RGBtoCMYconversion
TheconversionfromRGBtoCMYisdoneusingthismethod.
Consideryouhaveancolorimage,meansyouhavethreedifferentarraysofRED,GREEN
andBLUE.NowifyouwanttoconvertitintoCMY,hereswhatyouhavetodo.Youhaveto
subtract it by the maximum number of levels 1. Each matrix is subtracted and its
respectiveCMYmatrixisfilledwithresult.
YUV
YUV defines a color space in terms of one luma (Y) and two chrominance (UV)
components.TheYUVcolormodelisusedinthefollowingcompositecolorvideostandards.
NTSC(NationalTelevisionSystemCommittee)
PAL(PhaseAlternatingLine)
SECAM(Sequentialcouleuraamemoire,Frenchforsequentialcolorwithmemory)
168/179
10/26/2016
DIPQuickGuide
YCbCr
YCbCrcolormodelcontainsY,thelumacomponentandcbandcrarethebluediffernece
andreddifferencechromacomponents.
Itisnotanabsolutecolorspace.Itismainlyusedfordigitalsystems
ItscommonapplicationsincludeJPEGandMPEGcompression.
YUV is often used as the term for YCbCr, however they are totally different formats. The
maindifferencebetweenthesetwoisthattheformerisanalogwhilethelaterisdigital.
IntroductiontoJPEGCompression
169/179
10/26/2016
DIPQuickGuide
IntroductiontoJPEGCompression
In our last tutorial of image compression , we discuss some of the techniques used for
compression
WearegoingtodiscussJPEGcompressionwhichislossycompression,assomedataisloss
intheend.
Letsdiscussfirstwhatimagecompressionis.
Imagecompression
Imagecompressionisthemethodofdatacompressionondigitalimages.
Themainobjectiveintheimagecompressionis:
Storedatainanefficientform
Transmitdatainanefficientform
Imagecompressioncanbelossyorlossless.
JPEGcompression
JPEG stands for Joint photographic experts group. It is the first interanational standard in
imagecompression.Itiswidelyusedtoday.Itcouldbelossyaswellaslossless.Butthe
techniquewearegoingtodiscussheretodayislossycompressiontechnique.
Howjpegcompressionworks:
Firststepistodivideanimageintoblockswitheachhavingdimensionsof8x8.
Letsfortherecord,saythatthis8x8imagecontainsthefollowingvalues.
170/179
10/26/2016
DIPQuickGuide
Therangeofthepixelsintensitiesnowarefrom0to255.Wewillchangetherangefrom
128to127.
Subtracting128fromeachpixelvalueyieldspixelvaluefrom128to127.Aftersubtracting
128fromeachofthepixelvalue,wegotthefollowingresults.
Nowwewillcomputeusingthisformula.
171/179
10/26/2016
DIPQuickGuide
TheresultcomesfromthisisstoredinletssayA(j,k)matrix.
ThereisastandardmatrixthatisusedforcomputingJPEGcompression,whichisgivenbya
matrixcalledasLuminancematrix.
Thismatrixisgivenbelow
Applyingthefollowingformula
Wegotthisresultafterapplying.
172/179
10/26/2016
DIPQuickGuide
Now we will perform the real trick which is done in JPEG compression which is ZIGZAG
movement.Thezigzagsequencefortheabovematrixisshownbelow.Youhavetoperform
zigzaguntilyoufindallzeroesahead.Henceourimageisnowcompressed.
SummarizingJPEGcompression
ThefirststepistoconvertanimagetoYCbCrandjustpicktheYchannelandbreakinto8
x8blocks.Thenstartingfromthefirstblock,maptherangefrom128to127.Afterthat
you have to find the discrete fourier transform of the matrix. The result of this should be
quantized.Thelaststepistoapplyencodinginthezigzagmanneranddoittillyoufindall
zero.
Savethisonedimensionalarrayandyouaredone.
Note.Youhavetorepeatthisprocedureforalltheblockof8x8.
OpticalCharacterRecognition
173/179
10/26/2016
DIPQuickGuide
OpticalcharacterrecognitionisusuallyabbreviatedasOCR.Itincludesthemechanicaland
electricalconversionofscannedimagesofhandwritten,typewrittentextintomachinetext.
Itiscommonmethodofdigitizingprintedtextssothattheycanbeelectronicallysearched,
storedmorecompactly,displayedonline,andusedinmachineprocessessuchasmachine
translation,texttospeechandtextmining.
In recent years, OCR (Optical Character Recognition) technology has been applied
throughout the entire spectrum of industries, revolutionizing the document management
process. OCR has enabled scanned documents to become more than just image files,
turningintofullysearchabledocumentswithtextcontentthatisrecognizedbycomputers.
WiththehelpofOCR,peoplenolongerneedtomanuallyretypeimportantdocumentswhen
entering them into electronic databases. Instead, OCR extracts relevant information and
entersitautomatically.Theresultisaccurate,efficientinformationprocessinginlesstime.
Opticalcharacterrecognitionhasmultipleresearchareasbutthemostcommonareasare
asfollowing:
Banking:
he uses of OCR vary across different fields. One widely known application is in banking,
whereOCRisusedtoprocesscheckswithouthumaninvolvement.Acheckcanbeinserted
intoamachine,thewritingonitisscannedinstantly,andthecorrectamountofmoneyis
transferred. This technology has nearly been perfected for printed checks, and is fairly
accurate for handwritten checks as well, though it occasionally requires manual
confirmation.Overall,thisreduceswaittimesinmanybanks.
Blindandvisuallyimpairedpersons:
OneofthemajorfactorsinthebeginningofresearchbehindtheOCRisthatscientistwant
tomakeacomputerordevicewhichcouldreadbooktotheblindpeopleoutloud.Onthis
researchscientistmadeflatbedscannerwhichismostcommonlyknowntousasdocument
scanner.
Legaldepartment:
In the legal industry, there has also been a significant movement to digitize paper
documents. In order to save space and eliminate the need to sift through boxes of paper
files, documents are being scanned and entered into computer databases. OCR further
simplifies the process by making documents textsearchable, so that they are easier to
locateandworkwithonceinthedatabase.Legalprofessionalsnowhavefast,easyaccess
toahugelibraryofdocumentsinelectronicformat,whichtheycanfindsimplybytypingina
fewkeywords.
RetailIndustry:
174/179
10/26/2016
DIPQuickGuide
BarcoderecognitiontechnologyisalsorelatedtoOCR.Weseetheuseofthistechnologyin
ourcommondayuse.
OtherUses:
OCR is widely used in many other fields, including education, finance, and government
agencies. OCR has made countless texts available online, saving money for students and
allowingknowledgetobeshared.Invoiceimagingapplicationsareusedinmanybusinesses
to keep track of financial records and prevent a backlog of payments from piling up. In
government agencies and independent organizations, OCR simplifies data collection and
analysis,amongotherprocesses.Asthetechnologycontinuestodevelop,moreandmore
applications are found for OCR technology, including increased use of handwriting
recognition.
ComputerVisionandComputerGraphics
ComputerVision
Computer vision is concerned with modeling and replicating human vision using computer
software and hardware. Formally if we define computer vision then its definition would be
thatcomputervisionisadisciplinethatstudieshowtoreconstruct,interruptandunderstand
a3dscenefromits2dimagesintermsofthepropertiesofthestructurepresentinscene.
It needs knowledge from the following fields in order to understand and stimulate the
operationofhumanvisionsystem.
ComputerScience
ElectricalEngineering
Mathematics
Physiology
Biology
CognitiveScience
ComputerVisionHierarchy:
Computervisionisdividedintothreebasiccategoriesthatareasfollowing:
Lowlevelvision:includesprocessimageforfeatureextraction.
Intermediatelevelvision:includesobjectrecognitionand3DsceneInterpretation
175/179
10/26/2016
DIPQuickGuide
Highlevel vision: includes conceptual description of a scene like activity, intention and
behavior.
RelatedFields:
ComputerVisionoverlapssignificantlywiththefollowingfields:
ImageProcessing:itfocusesonimagemanipulation.
PatternRecognition:itstudiesvarioustechniquestoclassifypatterns.
Photogrammetry:itisconcernedwithobtainingaccuratemeasurementsfromimages.
ComputerVisionVsImageProcessing:
Image processing studies image to image transformation. The input and output of image
processingarebothimages.
Computer vision is the construction of explicit, meaningful descriptions of physical objects
from their image. The output of computer vision is a description or an interpretation of
structuresin3Dscene.
ExampleApplications:
Robotics
Medicine
Security
Transportation
IndustrialAutomation
RoboticsApplication:
Localizationdeterminerobotlocationautomatically
Navigation
Obstaclesavoidance
Assembly(peginhole,welding,painting)
Manipulation(e.g.PUMArobotmanipulator)
Human Robot Interaction (HRI): Intelligent robotics to interact with and serve
people
176/179
10/26/2016
DIPQuickGuide
MedicineApplication:
Classificationanddetection(e.g.lesionorcellsclassificationandtumordetection)
2D/3Dsegmentation
3Dhumanorganreconstruction(MRIorultrasound)
Visionguidedroboticssurgery
IndustrialAutomationApplication:
Industrialinspection(defectdetection)
Assembly
Barcodeandpackagelabelreading
Objectsorting
Documentunderstanding(e.g.OCR)
SecurityApplication:
Biometrics(iris,fingerprint,facerecognition)
Surveillancedetectingcertainsuspiciousactivitiesorbehaviors
TransportationApplication:
Autonomousvehicle
Safety,e.g.,drivervigilancemonitoring
ComputerGraphics
Computer graphics are graphics created using computers and the representation of image
data by a computer specifically with help from specialized graphic hardware and software.
Formally we can say that Computer graphics is creation, manipulation and storage of
geometricobjects(modeling)andtheirimages(Rendering).
The field of computer graphics developed with the emergence of computer graphics
hardware.Todaycomputergraphicsisuseinalmosteveryfield.Manypowerfultoolshave
been developed to visualize data. Computer graphics field become more popular when
companiesstartedusingitinvideogames.Todayitisamultibilliondollarindustryandmain
drivingforcebehindthecomputergraphicsdevelopment.Somecommonapplicationsareas
areasfollowing:
177/179
10/26/2016
DIPQuickGuide
ComputerAidedDesign(CAD)
PresentationGraphics
3dAnimation
Educationandtraining
GraphicalUserInterfaces
ComputerAidedDesign:
Usedindesignofbuildings,automobiles,aircraftandmanyotherproduct
Usetomakevirtualrealitysystem.
PresentationGraphics:
Commonlyusedtosummarizefinancial,statisticaldata
Usetogenerateslides
3dAnimation:
UsedheavilyinthemovieindustrybycompaniessuchasPixar,DresmsWorks
Toaddspecialeffectsingamesandmovies.
Educationandtraining:
Computergeneratedmodelsofphysicalsystems
MedicalVisualization
3DMRI
Dentalandbonescans
Stimulatorsfortrainingofpilotsetc.
GraphicalUserInterfaces:
It is used to make graphical user interfaces objects like buttons, icons and other
components
PreviousPage
NextPage
Advertisements
178/179
10/26/2016
DIPQuickGuide
Write for us
FAQ's
Helping
Contact
Copyright 2016. All Rights Reserved.

Enter email for newsletter
go
179/179

DIP Quick Guide

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

DIP Quick Guide

Hochgeladen von

Copyright:

Verfügbare Formate

10/26/2016

TheaboveimageofEinstein is an gray scale image. Means it is an image with 8 bits per

This method is known as histogram processing. We have discussed it in detail in previous

This method is known as transformations , in which we discussed different type of

Copyright 2016. All Rights Reserved.

Das könnte Ihnen auch gefallen