final project

| January 13, 2016

I have final project , I just I want from you to change the style of project and change introduction, and the most omportant thing I want from u to change conclusion, I do not want this professor to know that I take from my friend , any one does not have a lot review plz do not contact me , if u do not understand this is final and it is very important to me plz do not contact me .

thank u

Attachments:

GANNON UNIVERSITY

 

 

ELECTRICAL AND COMPUTER ENGINEERING DEPARTMENT

 

 

FALL2015

 

 

GECE 572: DIGITAL SIGNAL PROCESSING

 

 

 

FINGER PRINT RECOGNITION USING MINUTIAE BASED FEATURE

 

FINAL PROJECT

 

 

 

Prepared by

 

 

THADASINA PRUTHVIN REDDY

THADASIN002@gannon.edu

 

 

SALMAN SIDDIQUI

SIDDIQUI003@gannon.edu

 

 

 

 

Instructor:

Dr. Ram Sundaram

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table of contents

 

 

  1. Abstract
  2. Introduction
  3. Fingerprint matching
  4. Pre-processing stage
  5. Minutiaeextractionstage
  6. Post-processingstage
  7. Merits&Demerits
  8. Applications& futurescope
  9. Conclusions

10.References

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.Abstract

 

 

 

Nowadays,conventionalidentificationmethodssuchasdriver’slicense,passport,ATMcards andPINcodesdonotmeetthedemandsof thiswide scaleconnectivity.Automatedbiometricsin general,andautomatedfingerprintauthenticationinparticular,provideefficientsolutionsto thesemodernidentificationproblems.Fingerprintshavebeenusedformany centuriesasameans ofidentifyingpeople.Thefingerprintsofindividualareuniqueandarestayunchangedduring thelife time.Fingerprintmatchingtechniquescanbeplacedintotwocategories,minutiae-based andcorrelationbased.Minutiae-basedtechniques firstfind minutiae pointsandthen maptheir relativeplacementonthefinger.However,therearesomedifficultieswhenusingthisapproach. Itisdifficulttoextract theminutiaepointsaccurately whenthefingerprintisoflowqualitythe correlation-basedmethodisabletoovercome someof the difficultiesof theminutiae-based approach.However,ithassomeofitsownshortcomings.Correlation-basedtechniquesrequire the precise location ofa registrationpointand are affected byimagetranslation and rotation.

 

 

 

 

 

2.Introduction

 

 

Biometric recognitionrefers to the useof distinctivephysiological(e.g. fingerprint,palmprint, iris, face) and behavioral(e.g. gait, signature) characteristics, calledbiometricidentifiers for recognizing individuals.

 

 

Fingerprintrecognition is oneof theoldest andmostreliable biometric usedfor personal identification. Fingerprintrecognition has been used forover 100yearsnow and has comealong wayfrom tedious manualfingerprintmatching. The ancient procedureof matching fingerprints manuallywas extremelycumbersome and time-consuming and required skilled personnel.

 

 

Fingerskin is madeup of friction ridges andsweatpores all along theseridges.Friction ridges arecreated during fetal lifeand onlythe general shapeisgeneticallydefined. Thedistinguishing natureof physical characteristics ofaperson is dueto both the inherent individual genetic diversitywithin the humanpopulation as wellas the random processesaffecting the development ofthe embryo. Friction ridges remain the samethroughout one’s adultlife.They canreconstruct themselves even in caseof an injuryas long as theinjuryis not tooserious.

 

 

Fingerprints areoneof themostmaturebiometrictechnologies andareconsidered legitimate proofs of evidencein courts oflaw allover theworld.Inrecent times, more and more civilian and commercial applications areeitherusing or activelyconsidering using fingerprint-based identification becauseoftheavailabilityof inexpensive and compact solidstatescanners as well as its superior and proven matching performanceover other biometric technologies.

 

 

Some important terms related to fingerprintidentification systems areexplainedbelow:

 

  • FingerprintAcquisition:Howtoacquirefingerprintimagesandhowtorepresentthemin aproper machine-readable format.
  • Fingerprint Verification:To determinewhethertwo fingerprints arefrom the same finger.

 

  • Fingerprint Identification: To searchforaqueryfingerprintin a database.

 

  • Fingerprint Classification: To assign agiven fingerprintto one ofthe pre specified categoriesaccording to its geometric characteristics.

 

 

 

 

Incaseofbothfingerprintidentificationandfingerprintverification systems,ourtaskswill be broken up into 2 stages:

 

 

  1. Off-linephase:Severalfingerprintimagesofthefingerprintofapersontobeverifiedare firstcapturedandprocessedby afeatureextractionmodule;theextractedfeaturesare stored as templates in adatabaseforlater use.

 

 

  1. On-line phase: The individual to be verified gives his/her identity (in case of a verification system) and placeshis/her finger on the inklessfingerprintscanner, minutia pointsare extracted fromthecapturedfingerprintimage.Theseminutiaearethenfedtoa matchingmodule,whichmatchesthemagainsthis/herowntemplatesinthe database(in case of averificationsystem) or againstallthe usersinthe database(incase ofan identification system).

 

 

2.1    Whatisafingerprint?

 

Fingerprintsarethemostimportantpartinbiometricforhumanidentification.They areunique andpermanentfrombirthtodeath.So,fingerprintshavebeenusedfortheforensicapplication and personal identification.

 

 

Afingerprintiscollectionofmany ridgesandfurrows(Valleys).Thecontinuousdarkpattern flow infingerprintiscalledridgesandthelightarea betweenridgesiscalledfurrows. Fingerprint hassomeuniquepointsontheridge whichisknownasminutiae point.Herewe canconsider two maintypesof minutiaepointswhichare terminationpointandbifurcationpointasshownin Fig.1.Termination: wherea ridgeends and Bifurcation: whereridges splitinto two parts.

 

 

 

Figure1 MinutiaePoints (Termination, Bifurcation)

 

 

 

 

2.2     Fingerprint Recognition

 

 

 

The fingerprintrecognitionproblem canbe groupedintotwosub-domains:one isfingerprint verificationandthe other isfingerprintidentification.Inaddition,differentfromthe manual approachforfingerprintrecognitionby experts,thefingerprintrecognitionhereisreferredas AFRS(AutomaticFingerprintRecognitionSystem.Fingerprintverificationistoverify the authenticityofonepersonby hisfingerprint.Theuserprovideshisfingerprinttogetherwithhis identity informationlikehisIDnumber.Thefingerprintverificationsystemretrievesthe fingerprinttemplateaccordingtotheIDnumber andmatchesthetemplatewiththe real-time acquiredfingerprintfromtheuser.Usually itistheunderlyingdesignprincipleofAFAS (AutomaticFingerprintAuthentication System).

 

 

 

Figure 2. Generalarchitecture ofafingerprintverification system

 

 

 

Fingerprintidentificationistospecify oneperson’sidentity by hisfingerprint(s).Without knowledge of theperson’sidentity,thefingerprintidentificationsystemtries tomatchhis fingerprint(s)withthoseinthewhole fingerprintdatabase.Itisespeciallyusefulforcriminal investigationcases.Anditisthedesignprinciple ofAFIS(AutomaticFingerprintIdentification System).However,allfingerprintrecognitionproblems,either verificationoridentification, are ultimatelybasedonawell-definedrepresentationofafingerprint.Aslongastherepresentation offingerprintsremainstheuniquenessandkeepssimple,thefingerprintmatching,eitherforthe

1-to-Iverification caseor 1-to-m identification case,is straightforward andeasy.

 

 

 

 

2.3    Techniques for Fingerprint Recognition

 

 

 

1) MinutiaeExtractionbasedTechniques:Mostlyacceptedfingerscantechnologyisbasedon Minutiae. Minutiae basedtechniques produce the fingerprint by  its local features, like terminationandbifurcation. Whenminutiae pointsmatchbetweentwofingerprintssofingerprint are match.Thisapproachhasbeengenuinely studied,anditisthebackboneofthecurrent available fingerprintrecognition products.

 

 

2) PatternMatchingorRidgeFeaturebasedTechniques:Featureextractionareestablishedon series ofridges as averse  to different points which design the basis of  pattern  matching techniques over Minutiae Extraction.Minutiae pointscanbechange bywearandtearandthe main drawback arethat these areacute to proper adjustment of finger and need largestorage.

 

 

3) CorrelationbasedTechniques:Correlationbasedtechniqueisusedtomatchtwofingerprints, thefingerprintareadjustedandcomputedthecorrelationforeachcorrespondingpixel.They can matchridgeshapes,breaks,etc.Maindisadvantages ofthismethodare itscomputational complication and less toleranceto non-linear distortion and contrast variation.

 

 

4)Image based Techniques:Thistechniqueattempttodomatchingwhichbasedontheglobal featuresofanallfingerprintimages.Itisanadvanceandnewly developsmethodforfingerprint recognition

 

 

 

 

 

 

3.Fingerprintmatching

 

 

Thematchingoffingerprintisachieved by someimageprocessingsteps.Thesestepcaneasily beunderstandbythealgorithm below:

Input:  TwoGray-scale Fingerprintimage.

 

Output:Verifythe fingerprintimageusing minutiaematching.

 

 

 

Step 1: Enhancement ofInputImagei.e. fingerprint imageusing Histogram equalization. Step 2:  Binarized the enhanced fingerprintimage.

Step 3: Selection ofROI(Region ofInterest) in binarized image.

 

Step 4: Thinning ofthe Region ofInterest as the part of fingerprintimage. Step 5:  Minutiaepointsareextracted from image.

Step 6: Comparison and matching ofone fingerprintto another fingerprint.

 

Step 7: Match the minutiaepoints of two images arecomputed.IfMinutiaepoints arematched in  both  images  so  fingerprint matching  score are1 and  if  it  is  not  matched then matching scoreare0 theyaremismatched.

 

 

Figure3. FingerprintMatchingblockdiagram

 

 

 

Theoverallimplementationofalgorithmmay alsoexpress byusingblockdiagram,asshown above.Thisblockdiagramissub dividedaspre-processingstage,minutiaeextractionstageand post-processing stage

 

 

 

 

 

4.Pre-processingstage

 

 

4.1    Image Acquisition

 

 

 

Thefirststageofanyvisionsystemwhetherforidentificationorverificationis  theimage acquisitionstage.Nowadays,the automatedfingerprintverificationsystemsuse live-scandigital imagesoffingerprintsacquiredfroma fingerprintsensor.Thesesensorsare basedonoptical, capacitance, ultrasonic, thermal and other imagingtechnologies.

 

 

  1. Optical Sensors: These aretheoldest andmostwidelyused technology.In most devices, a chargedcoupled device(CCD) converts the imageof the fingerprint, with dark ridges and light valleys, into adigital signal. Theyare fairlyinexpensive andcan provide resolutions up to 500 dpi. Mostsensors arebased on FTIR (Frustrated TotalInternal Reflection) technique

to acquirethe image.In thisscheme, asourceilluminates the fingerprintthroughoneside

 

 

 

 

Figure4 :(a) General schematicforan FTIR based optical sensor (b)Schematic ofacapacitive sensor

 

 

of the prismasshown(Figure 4).Duetointernalreflectionphenomenon,mostof thelightis reflectedbacktotheothersidewhereitisrecordedbyaCCDcamera.However,inregions where thefingerprint surface comes in contact with the prism,  the light is diffused in all directionsandthereforedoesnotreachthesensor resultingindarkregions.The qualityof the imagedependsonwhetherthefingerprintisdryorwet.Anotherproblemfacedbyoptical sensorsistheresidualpatternsleftby thepreviousfingers.Furthermoreithasbeenshownthat fakefingersareableto  foolmostcommercialsensors.Opticalsensorsare alsoamongthe bulkiest sensor dueto theoptics involved.

 

 

 

 

2.CapacitiveSensors:Thesiliconsensoractsasoneplateofacapacitor,andthefingeras another other.The capacitance betweenthesensingplate andthefingerdependsinverselyasthe distancebetween them. Sincethe ridgesare closer,theycorrespond to increased

capacitanceandthevalleyscorrespondtosmallercapacitance.Thisvariationisconvertedintoan

 

8-bitgray scaledigitalimage.Mostoftheelectronicdevicesfeaturingfingerprintauthentication use this form of solid statesensorsdue toits compactness. However, sensorsthat aresmallerthan

0.5”x0.5” arenot useful sinceitreduces the accuracyrecognition.

 

 

 

  1. Ultra-sound Sensors:Ultrasound technologyis perhaps themostaccurate ofthe fingerprint sensing technologies.It uses ultrasound waves andmeasures the distancebased on theimpedance ofthefinger,theplate,andair.Thesesensorsarecapableofveryhighresolution.Sensorswith

1000dpiormorearealready available(www.ultra-scan.com).However,thesesensorstendtobe verybulkyandcontainmovingpartsmakingthemsuitableonly forlawenforcementandaccess control applications.

 

 

  1. Thermal Sensors:Thesesensors aremadeup of pyro-electric materialswhoseproperties change withtemperature. These areusually manufactured in the form of strips .As the fingerprintsisswipedacrossthesensor,thereisdifferentialconductionofheatbetweenthe ridgesandvalleys(sinceskinconductsheatbetterthantheairinthevalleys)thatismeasuredby thesensor.Fullsizethermalsensorsarenotpracticalsinceskinreachesthermalequilibriumvery quicklyonceplacedonthesensorleadingtolossofsignal.Thiswouldrequireustoconstantly keepthesensoratahigherorlowertemperaturemakingitveryenergyinefficient.Thesweeping actionpreventsthefingerfromreachingthermalequilibriumleadingtogoodcontrastimages. However, sincethe sensor can acquire only small strips at a time, a sophisticated image registration and reconstruction scheme is requiredto constructthe whole image from the strips.

 

 

One of themostessentialcharacteristicsof a digitalfingerprintimage isitsresolutionwhich indicatesthenumberofdotsorpixelsperinch (ppi).The minimumresolutionthatallowsthe feature extraction algorithms to locate minutiaeis 250 to 300ppi.

 

 

 

 

4.2    Image Enhancement

 

 

 

Fingerprintimage enhancement is to makethe image clearer foreasyfurther operations.

 

Theperformanceofminutiaeextractionalgorithmsandotherfingerprintrecognitiontechniques relies  heavilyon thequalityof theinput  fingerprint  images. Afingerprint  imageis  firstly enhancedbeforethefeaturescontainedinitcouldbedetectedorextracted.Awellenhanced image willprovide a clear separation between the valid and spurious features. Since the fingerprintimages acquired from sensors or othermedia arenotassured with perfect quality. Howeverthefingerprintimagesobtainedareusuallypoorduetoelementsthatcorrode theclarity oftheridgeelements.Thisleadstoproblemsinminutiaeextraction.Spuriousfeaturesarethose minutiaepointsthatarecreatedduetonoiseorartifactsandtheyarenotactuallypartofthe fingerprint.

Inanidealfingerprintimage,ridgesandvalleysalternateandflowinalocally constant direction.Thus,imageenhancementtechniquesareemployedtoreducethenoiseandenhance thedefinitionofridgesagainstvalleys.Inordertoensure goodperformance oftheridge and minutiaeextractionalgorithmsinpoorquality fingerprintimages,anenhancementalgorithmto improvetheclarity oftheridgestructureisnecessary.Enhancementmethods,forincreasingthe contrastbetweenridgesandfurrowsandforconnectingthe false brokenpointsof ridgesdue to insufficient amount of ink areveryuseful to keepahigheraccuracyto fingerprintrecognition.

 

 

HistogramEqualization

 

 

 

Itisa methodfor enhance thefingerprintimage. Fingerprintimage enhancementistocreate clearerforeasy otheroperations.Histogramequalizationistoextendthepixelvalueofanimage soastoincrease the perceptionalinformation.The histogramofaoriginalfingerprintimage has thebimodaltypethehistogramafterthehistogramequalizationoccupiesalltherangefrom0to

255 and the visualization effect is enhanced.

 

 

 

In MATLAB histogramequalization is donebyusing MATLAB function.

 

histeq(image_file_name);

 

 

 

 

Below, the figureshowsthe original imagehistogram and histogram afterequalization operation.

 

 

 

 

 

 

4.3    Binarization

 

 

AFingerprint-Image-Binarizationtransformsan8-bitgrayimagetoa1-bitbinarizedimage. Mostminutiaeextractionalgorithmsoperateonbinary imageswherethereareonlytwolevelsof interest:0-value holds for ridgesand1-value for furrows.And after thebinarizationoperation ridges arehighlighted with black color andfurrows arehighlighted with white color.

 

Anadaptive binarizationmethodisachievedtobinarize thefingerprintimage.Inthismethod image issplitintoblocksof 16x16pixels.Apixelvalueisset1ifitsvalueisgreater thanthe mean intensityvalueof the acceptedblock to which the pixelbelongs.

 

 

 

 

 

 

 

 

 

 

 

4.4    Image segmentation

 

 

Thisisa segmentationtechnique. The mainmotive of thesegmentationistomake theimage simplerwhichcanberepresentingveryeasily andtomakeimagemeaningfulthatwillbeeasyto analyze.Generally ROI(RegionofInterest)isveryusefulforanalyzeafingerprintimage.Itisa subsetof animage ora datasetanalyze foraparticular purpose. Whentheimage area has ineffective ridges andfurrows so firstlyitmadewider and larger in alldirections.

 

Therearetworegionsthatdescribeanyfingerprintimage;namely theforegroundregionandthe backgroundregion. Theforegroundregions arethe regionscontaining the ridges and valleys.The ridgesare theraisedanddarkregionsof afingerprintimage whilethe valleysarethe lowand white regionsbetweenthe ridges.The foregroundregionsoften referredtoasthe Regionof Interest(ROI).Thebackgroundregionsaremostly theoutsideregionswherethenoises introduced into the imageduring enrolment aremostlyfound

 

RegionofInterest (ROI)

 

 

To extraction oftheROIis performed in twosteps: First, block directionestimation and direction varietycheck;second, used someMorphologicalmethods.

 

Two typesof morphologicalmethodsareavailable i.e. OPENandCLOSE.The OPENoperation can  enlargetheimages  and  eliminatebackgroundnoise.AndCLOSE  operationcan  shrink images and eliminate smallcavities.

 

bwmorph (x,’close’); bwmorph(y,’open’);

 

 

 

 

 

 

 

5.Minutiaeextractionstage

 

 

 

Aftertheenhancementofthefingerprintimage,theimageisready forminutiaeextraction.For proper extraction,however, a thinningalgorithmisappliedtothe enhancedimage.Itproducesa skeletonized representation of theimage.

 

 

5.1    Thinning

 

 

 

Thinningisa morphologicaloperationthatisused toremove selectedforegroundpixels from binaryimages.Itisusedtoeliminatetheredundantpixelsofridgestilltheridgesarejustone pixelwide.Thinningisnormally onlyappliedtobinary images,andproducesanotherbinary imageasoutput.Itisthefinalsteppriortominutiaeextraction.Allthepixelsontheboundaries offoregroundregionsthathaveatleastonebackgroundneighboraretaken.Any pointthathas morethanoneforegroundneighborisdeletedaslongasdoingsodoesnotlocally disconnectthe region containing that pixel.

This is donebyusing theMATLABthinning function thatis:-

 

 

  • bwmorph(binaryImage,’thin’,Inf)

 

 

 

 

Thenthethinnedimageisfilteredby usingthefollowingthreeMATLABfunctions.Thisare some H is breaks, isolated points andspikes.

 

  • bwmorph(binaryImage,’hbreak’,k)

 

  • bwmorph(binaryImage,’clean’,k)

 

  • bwmorph(binaryImage,’spur’,k)

 

 

 

 

The conditions forbetterthinning result:

 

 

  1. a) Each ridgeshould bethinned to itsce b) Noise and singularpixelsshould be removed.
  2. c) No furtherremoval ofpixels should be possible after accomplish of thinning process

 

 

 

 

 

 

5.2    Minutiae Marking

 

 

The methodextractstheminutiae fromtheenhancedimage.Thismethodextractstheridge endingsandbifurcationsfromtheskeletonimageby examiningthelocalneighbourhoodofeach ridgepixelusinga3×3window.Themethodusedforminutiaeextractionisthecrossing number (CN)method.Thismethodinvolvestheuseoftheskeletonimage where theridgeflow patterniseight-connected.Theminutiaeareextractedby scanningthelocalneighbourhoodof eachridgepixelintheimageusinga3×3window.CNisdefinedas halfthesumofthe differences between thepairs of adjacent pixel.

 

 

 

 

 

 

CN=0.5i=1Σ8(Pi-Pi+1)

 

Theridgepixelcanbedividedintobifurcation,ridgeendingandnon-minutiaepointbasedonit. A ridgeending point  has onlyone neighbor, a  bifurcation point possesses more than two neighbors,andanormalridge pixelhastwoneighbors. A CNvalueof zero referstoanisolated point,value of one toaridge ending,twotoa continuingridge point,three toa bifurcationpoint anda CN of four meansa crossingpoint.Minutiae detectionina fingerprintskeletonis implemented by  scanning thinned fingerprint and counting the crossing number. Thus the minutiaepoints can be extracted.

Cn{p} =1 RidgeEnding Cn{p} =3 RidgeBifurcation

 

 

 

 

 

 

 

 

Inthe proposedmethod,the minutiaepoint’slocationsandtheir considereddirectionfromthe 8 directions(N,S,W,E,NE,NW,SE,SW)arerecordedthenthey usedtoconstructthedatabase depending of thenumberof recorded minutiaepoint and their direction.

SupposePis the checkedpointand P1-P8 areneighbourhood pixels

 

IfCN =3 then

 

If P1 and P3 and P7 =1 then direction =W Else ifP1 and P3 and P5 =1 then direction =S Else ifP1 and P7 and P5 =then direction =N Else ifP3 and P5 and P7 =1 then direction =E Else ifP4 and P3 and P5 = 1 then direction =SE Else ifP3 and P2 and P1 =1 then direction =SW Else ifP3 and P5 and P6 =1 then direction =NE

Else ifP4 and P8 and P5 =1 then direction =NW End if

IfCN =1 then

 

If P1 =1 then direction =W If P1 =1 then direction =W If P3 = 1 then direction =S If P7 =1 then direction =N If P5 =1 then direction =E

If P4 =1 then direction = SE If P2 =1 then direction = SW If P6 =1 then direction =NE If P8 =1 then direction =NW

 

 

 

 

 

6.Post-processingstage

 

 

 

 

6.1    Minutiae Matching

 

 

Whenallminutiaepointsoftwofingerprintimagesareextractedinselectedregionofinterest. Now, minutiae matchingareperformed forverification. Basically, minutiae

 

Matching is aprocesswhich completed in two steps:

 

 

1) Find Total MinutiaPoints:  This step is used to calculate the total numberof Ridge and Bifurcationpointsseparately.Anditcomparesthecalculatedvaluewiththeoriginalimage values.

2) FindLocationof Minutiae Points:Itworksonthebasisof Minutia Markingprocess.Simply, whenminutia pointsmarkedontheimage italsostore thelocationof thepoint.Thisstored informationitusedtocomparetwodifferentimagesatverificationprocess.Ifboththe images belongto the  same  person  then  the location of  ridge/bifurcation will match. Otherwise matching offingerprintimages unsuccessful.

 

 

 

6.2     Remove False Minutiae

 

Infingerprintrecognition,thegoalistoo abletodetecttheminutiaepointandtoreducethefalse minutiaeinthefingerprintimage.Inordertoremovefalseminutia,thereare a fewprocessthat need to be going throughwhich areminutiamarking and false minutia removal.

Theprocedures in removing false minutia are:

 

  1. Ifthe distancebetween onebifurcation and onetermination is less than D and the two minutiae arein thesame ridge (ml case). Remove both ofthem. WhereD isthe averageinter-ridgewidthrepresenting the averagedistancebetween twoparallel neighboring ridges.
  2. If thedistancebetween two bifurcations is less than D and theyarein thesame ridge, remove thetwo bifurcations.
  3. Iftwo terminations arewithin a distanceD and their directions are coincident with asmall angle variation. And theysufficethecondition that no anyother

termination is located between the two terminations. Then the two terminations are

 

 

 

 

 

7.Merits&Demerits

 

 

 

 

Advantages

 

 

  • Physical attributes aremuch tougherto be faked thanID cards.
  • Fingerprints can’t be guessed unlikepasswords.
  • Fingerprints can’t be misplaced unlikeacard.
  • Fingerprints can’t be forgotten unlikepasswords.
  • Sudden enhancement in the current securitylevel.
  • Less securityconcerns leads to increased productivity.

 

 

Disadvantages

 

 

  • It can bedeceived byapictureoramold of fingerusingGelatin.
  • Fingerprints if stolen, can beagreat threat toSecurityand intellectual property.
  • Requires a verylargedatabaseoffingerprints.
  • Some ofthe employeesmayfind it uncomfortabletoHavetheirfingerprint stored with the employer.

 

 

 

 

 

8.Applications& futurescope

 

 

 

 

 

Applications

 

 

  1. Financial services(e.g. ATM )

 

  1. Immigration &border control (e.g. points ofentry declared for   frequent travelers, passport and visacases )
  2. Social services (e.g. fraudpreventationin entitlement programmers)

 

  1. Health care(e.g. securitymeasure for privacyormedical records)

 

  1. Physical accesscontrol (e.g. at institutional, government & residential establishment)

 

  1. Time &attendance(e.g.replacement oftime punch card)

 

7.ComputerSecurity (e.g.personalcomputeraccess,networkaccess,Internetuse,e- commerce,e-mail, encryption)

  1. Telecommunications(e.g.mobilephones,callcentertechnology,phonecards,televised shopping)
  2. Lawenforcement(e.g.criminalinvestigation,nationalID,drivinglicense,rehabilitation institutions/prison, home confinement, smallgun)

 

 

Furtherworks which can be carried out includefollowing.

 

 

  1. Toperformstatisticalexperimentusedinthisprojectonalargersamplesize&aconduct a fullanalysisofobserved result.

2.Animplementationof a smarter matchingalgorithmshouldbe abletoimprove the verification&identification process.

  1. Issueneedtobeaddressedinthesystematicwayindevelopingafoolprooffingerprint basedidentification systemforawidescaledevelopmente.g.encryptionsecurity of fingerprinttemplatedetection of force fingers, privacyconcernetc.
  2. Implementation ofon-line fingerprintverification&Identification system using biometric device.

 

 

 

 

 

9.Conclusion

 

 

Thisseminarhasconcentratedonfingerprintbasedbiometric identification&verification systems.Theprimaryfocusissubsequentextractionofminutiaeby directgrayscaleimage extractiontechnique .Therearetwoimportant operationsinpre-processingstage asHistogram Equalization,and Selectionof ROI.These twooperationsmake thisalgorithmefficient.The HistogramEqualizationenhancedthequality ofInput-image,whichactually helptoproduce accuratecalculation.ThisresearchconcludesthattheFingerprintVerificationispossibleeven thequality ofthefingerprintimagegotaffected.TheROIbasedapproachreducestheprocessing timeofalgorithm  byworking  on  segment  not  thecompleteimage,  whichmeans  it  makes fingerprintmatchingfaster.Theverificationisdonefor selectedregionthatauthenticatethe pattern.Theliteratureofthistechniqueisdeeply studiedandexperimentally executedin MATLAB.

 

 

 

 

 

10.References

 

 

 

 

  1. Raymond Thai, ‘FingerprintImageEnhancementand Minutiae-Extraction,”

ThesissubmittedtoSchoolofComputerScienceandSoftwareEngineering,Universityof

Western Australia

 

  1. AKJain, A. Ross, and S.Prabhakar, FingerprintMatching Using Minutiae

and TextureFeatures ,Proc. ofInternational ConferenceonImageProcessing,2001

 

  1. DigitalImageProcessingbyRafaelC.GonzalezandRichardE.Woods,PearsonEducation,

2003

 

  1. Digital ImageProcessing using MATLAB: Rafael C. Gonzalez, RichardE. Woods 2nd

Edition, 2009

  1. FingerprintImageEnhancement and MinutiaeExtraction byRaymond Thai 2002

 

 

  1. Online FingerprintVerification bySharatChikkerurCUBS, UniversityofBuffalo

 

Get a 5 % discount on an order above $ 150
Use the following coupon code :
2018DISC
Project: Experimental determination of a Servo-Motor State Space Model
final project

Category: Homework Help

Our Services:
Order a customized paper today!