Basictypes
Therearethreebasicshapesoffingerprintlines-whorl,archandloop.Itisthesubcutaneoustissueoppositetothetopofthefingerbelly.Thedifferentpressingdirectioncreatesthisdifferentshape.Studieshaveshownthatifsomeone’sfingeristallandround,thepatternoftheirfingerprintswillbespiral.Scientistshavebeenabletoreproducethemorecommonfingerprintsthroughthemodel,andcanalsorepeattheformationprocessofthelesscomplicatedrarefingerprints.
Fingerprintstructure
Fingerprints,alsocalledhandprints,havebroadandnarrowmeanings:fingerprintsinanarrowsenserefertothepatternofmastoidlinesonthepalmsurfaceofthefirstsectionofahumanfinger;Inabroadsense,fingerprintsincludefingerprints,knuckleprintsandpalmprints.Thereisaliteraldifferencebetweenfingerprintsandfingerprints,thatis,fingerprintsrefertothepapillarylinepatternonthepalmsurfaceofthefirstfinger,andfingerprintsarethemarksleftbythispapillarylinepattern.However,injudicialpractice,itiscustomarythatfingerprintsandTheconceptoffingerprintsisuniversal.
Theimprintsaremainlyduetothepresenceofalargenumberofsweatglandsandsebaceousglandsontheskinofhumanfingersandpalms(thinkofthesweatinyourpalmswhenyouarenervousorexcited).Aslongaslifeactivitiesexist,Sweatandsebumarecontinuouslydischarged.It'sabitlikeanatomicsealthathasinkconstantlypermeatingthesurfaceoftheprintedtext.Therefore,aslongasthefingerorpalmtouchesthesurfaceoftheobject,itwillautomaticallyleaveamarklikeanatomicseal.Ofcourse,thisismainlythereasonwhythefingersandpalmsthemselvescanleavefingerprints.Ifthefingersandpalmsarestuckwithotherliquid-likesubstances,suchasoilsontheheadandface(thisisthe"inkpad"youusemostoften),blood,andinktoholdfingerprints,theprincipleofleavingfingerprintsismorelikeanormallid.seal.
Fingerprintforensics,includingfingerprintsearchanddiscovery.Thescopeoffingerprintsearch:(1)thecriminalactivitycenter;(2)theimportandexportofthesceneanditssurroundings;(3)theitemsthatthecriminalsmayhavetouched;(4)thevariouslethalweaponsandobjectsleftbythecriminalsonthescene.
Andtheircomplexityissufficienttoprovideenoughfeaturesforidentification.Inadditiontobeingunique,fingerprintsarealsohereditaryandimmutable.Ithasnotbeenfoundthatdifferentpeoplehavethesamefingerprints,soeachperson'sfingerprintsarealsodifferent.Becausefingerprintsareuniquetoeveryone,fingerprintsleftbycriminalsatthesceneofcrimeshavebecomeimportantcluesforthepolicetopursuesuspectsinthepastfewhundredyears.Nowadays,fingerprintidentificationmethodshavebeencomputerized,makingtheidentificationprocessfasterandmoreaccurate.
Becauseeveryonehasdifferentgeneticgenes,fingerprintsarealsodifferent.However,althoughtheformationoffingerprintsismainlyaffectedbygenetics,therearealsoenvironmentalfactors(5%).Whenthefetusdevelopsinthemother'sbodyforthreetofourmonths,thefingerprintsarealreadyformed,andthefingerprintsofchildrenwillslightlychangeduringthegrowthperioduntiladolescence.Theshapeisonlysetataround14yearsold.Intheprocessofskindevelopment,althoughtheepidermis,dermis,andstromallayersallgrowtogether,thesoftsubcutaneoustissuegrowsfasterthantherelativelyhardepidermis.Somepeoplesaythatfingerprintswillchangeafterbonemarrowtransplantation,whichiswrong.Unlessitisskingraftingordamagethatreachesthebasallayer,fingerprintswillnotchange.
Formthefoundation
Duringthedevelopmentoftheskin,althoughtheepidermis,dermis,andstromallayersareallgrowingtogether,thesoftsubcutaneoustissuegrowsfasterthanthehardepidermis,soitwillTheepidermisproducesasteadystreamoftoppressure,whichforcestheslower-growingepidermistoshrinkandcollapsetowardtheinnerlayeroftissues,graduallybendandwrinkle,soastoreducethepressureexertedbythesubcutaneoustissuesonit.Inthisway,ontheonehand,itisforcedtoattackupwards,andontheotherhand,itisforcedtowithdraw,causingtheskintogrowtortuously,pitsanduneven,forminglines.Thisprocessofbendingandwrinklingfluctuateswithchangesintheupperpressuregeneratedbytheinnertissue,formingunevenridgesorfolds,untilthedevelopmentprocessisstopped,andfinallyshapedintoafingerprintthatwillnotchangetodeath.
Geneticproblems
Canfingerprintsbeinherited?
Althoughtherearenotwoidenticalfingerprintsintheworld,thefingerprintsofidenticaltwinsarerelativelysimilar.Moreover,thepatternsofdifferentfingersofthesamepersonalsohavesimilarities.Thestudyfoundthattherearegenderdifferencesandethnicdifferencesinfingerprinttypeswithhighoccurrencerates.Forexample,inthefingerprintsofChineseandJapanese,theoccurrencerateofbucketpatternandskippatternisroughlythesame,accountingformorethan90%ofthetotal;inthefingerprintsofEuropeans,theoccurrencerateofskippatternishigher;inthefingerprintsofAmericans,Theappearancerateofbowpatternishigher.
Becausethefingerprintsofidenticaltwinsorthesameracearesimilar,itcanalsobesaidthatfingerprintscanbe"inherited".However,theformationoffingerprintsisnotonlyaffectedbygeneticgenes,butalsobyenvironmentalfactors.ProfessorKazuInoueofHokkaidoUniversityinJapanhasbeenengagedinanatomicalresearchforalongtime.Hesaid:"Thedetailsofhumanbodyalsovaryfrompersontoperson.Becausefingerprintsareveryconvenienttouse,theyarewidelyusedtoidentifyidentity."
Howarefingerprintsformed?
Whenthefetusdevelopsto4months,ithasalreadyformedafingerprint.Beforethat,about10weeksofdevelopment,fingertipsandotherpartswilltemporarilyformalargeball-shapedbulge—justlikethepadofacat'spaw,whichplaysadecisiveroleintheformationoffingerprints.
Whenthebulgebeginstoshrinkandcollapse,wrinklesbegintoappearatthejunctionoftheepidermisandthedermis.Thisisthebudding"fingerprintmold".Thecellsgeneratedbythefingerprintmoldsqueezehardtothesurface,andwhenthefetusdevelopsto4months,itwillformfingerprintsontheepidermis.Researchsuggeststhatfingerprintsaredevelopedfromdenselyarranged"molds"onthebulgesurface.Therefore,theshapeandsizeoftheoriginalbulgedeterminestheshapeofthefingerprint.
Usesoffingerprints
Asweallknow,fingerprintshavethecharacteristicsof"differentandunchangingforlife".Longago,peopleusedhandprintsonpaperorwoodenboardstoidentifytheiridentity.Fingerprintshavebeenwidelyusedinareassuchasentryinspectionsandcriminalsearches.Fingerprintsarelinesformedbylinearlyarrangedprotrusionsanddepressionsontheepidermis,"akindofskinpattern".Humanhandsandfeetarecoveredwithskinlinesontheinside.Fingerprintsarewidelyusedinareassuchassearchingforcriminalstoidentifyidentities.Asweallknow,theBabyloniansandChinausedfingerprintstoverifytheidentityofapersonalongtimeago.
Usingfingerprintstoidentifyidentity,thisisbecausefingerprintsmeetthefollowingtwoconditions:
Firstofall,fingerprintsarenotthesame,andtherearenotwoidenticalfingerprints.Althoughidenticaltwinshaveahighdegreeofsimilarityintheirpatterns,theirdetailedfeaturesarenotexactlythesame,buttherearecertaindifferences(seeFigure1forfeaturepoints).Inaddition,thefingerprintsofdifferentfingersofapersonarealsodifferent.
Secondly,aperson’sfingerprintisinprinciplethesameforlife.Whenachildgrowsup,thefingerprintisonlyenlargedandthickened,anditspattern,numberoflinesandothercharacteristicsremainunchanged.
Figure2showsthethreetypesoffingerprints(skippattern,bucketpattern,bowpattern),andshowsasubcategoryofbucketpatternandbowpattern.Thepictureontherightshowstheinternalstructureoftheskin.Althoughthereareothertactileorgans,onlythepartsinvolvedinthetextareshownhere.
Thedermisislocatedinthelowerlayeroftheepidermis,andthejunctionbetweentheepidermisandthedermisisunevenandintricate.Theseconcavitiesandconvexitiesare"molds"andfinallyformfingerprintpatterns.Eveniftheepidermisiswornoff,aslongastheinnerdermislayerisnotdamaged,thesamefingerprintscanstillgrowaftertheinjury.Theclarityoffingerprintswillgraduallyblurwithage.In1880,Fulders,aBritishmissionaryanddoctorlivinginJapan,publishedthefirststudyonfingerprintsintheBritishacademicjournal"Nature",expoundingforthefirsttimethatfingerprintrecognitionisusedincriminalinvestigationandotherfields.Theapplicationof,createdaprecedentformodernfingerprintresearch,whichwasintroducedtoAsiain1900andwasputintopractice.
Peoplehavealwaysbelievedthatfingerprintscannotonlyhelppolicesolvecrimes,butalsoenhanceskinfriction.However,afterscientistsmeasuredtheinfluenceoffingerprintsonfriction,theycameupwithdifferentresults.
Thescientistsfirstaskedthevolunteerstopresstheirfingersontheglasstocalculatetheaveragefrictioncausedbythefingerprintsatthistime;then,theyaskedthevolunteerstograduallyincreasethepressureonthefingersandperformcalculations.Itturnedoutthatthefrictiondidnotincreaseasexpected.Furthermicroscopicexaminationrevealedthatthefingerprintslookedlikegulliesunderthemicroscope,withgapsbetweenthem.Comparedwithaperfectlysmoothfingersurface,thecontactareawasreducedbyabout1/3.Thisisabitsimilartorubber,andthefrictionvarieswiththecontactarea.Basedonthis,scientistsbelievethatfingerprintsactuallyreducefriction,makingtheskineasiertostretchanddeform,sothatitcanavoidskindamage.
Observationmethod
Accordingtothedifferentlegacymethods,fingerprintscanbedividedintothreecategories:
Thefirstcategoryisthepatentprint,whichisthetargetprint.Linescanbeseenbysight.Forexample,hand-dippedpaint,blood,inkandotheritemsaretransferred,usuallyprintedonthefingerprintcardtobecomethebasicdata;
Thesecondtypeisplasticprint,whichreferstothesoftSubstances,suchasfingerprintsimprintedoncandlesandclaybyhandcontact;
Thethirdtypeislatentprint,whichisafingerprintthatistransferredbythebody’snaturalsecretionssuchassweat.Thelinesarenoteasytofindvisually,andarethemostcommonfingerprintsinthecrimescene.Latentfingerprintsareoftenleftafterthefingersfirsttouchgrease,sweatordust,andthentouchacleansurface.Althoughthesefingerprintscannotbeseenwiththenakedeye,theycanbeprocessedbyspecialmethodsandusingsomespecialchemicalreagents.Revealtheselatentfingerprints.
Physicalmethod
Iffingerprintsareleftonthesurfaceofnon-absorbentmaterialssuchasmetals,plastics,glass,tiles,etc.,physicalmethodscanusuallybeused.Fingerprintsappear.
(1)Powdermethod,choosepowderwithlargecolorcontrast,andsprinkleitontoextractcompletefingerprints;
(2)Magneticpowdermethod,usefineironpowderparticleswithmagnetsAsabrush,swipebackandforthtorevealfingerprints.
(3)Lasermethod.Withthedevelopmentoflasertechnology,ourcountryuseslasertodisplayfingerprints.Thedisplaydeviceusesanargonionlaser.Laserscandisplayfingerprintsbecausethereisalwaysalayerofsweatandfattyacidsonthesurfaceofhumanfingers,whichleavesinconspicuousfingerprintsaftertouchingobjects;withlaserlight,sweat,fattyacids,etc.willproducecolorfluorescence,andthefingerprintswillbeclear..Useadedicatedfingerprintcameratotakeclearfingerprintphotos.Afterthephotoisenlarged,itbringsalotofconveniencetotheappraisalwork.
Chemicalmethod
Iffingerprintsareleftonthesurfaceofabsorbentarticlessuchaspaper,cards,leather,wood,etc.,theymustbechemicallytreatedbeforetheycanberevealedinthelaboratory.
Iodinefumigationmethod-theuseofiodinecrystalstoheattogeneratesteam,afteritreactswiththegreaseofthefingerprintresidue,yellow-brownfingerprintswillappear,whichmustbetakenimmediatelyorfixedbychemicalmethods;
Ninhydrinmethod-spraythereagentonthespecimen,andafterreactingwiththeaminoacidsinbodysecretions,itwillshowapurplefingerprint;
Silvernitratemethod-nitricacidAfterthesilversolutionreactswiththesodiumchlorideintheresidualsweat,itwillproduceblackfingerprintsinthesun;
Fluorescencereagentmethod—fluoresceinando-phthalaldehydequicklyinteractwiththeproteinorfingerprintresiduesTheactionofaminoacidsproduceshighlyfluorescentfingerprints.Thisreagentcanbeusedonthesurfaceofcoloredobjects.
Three-secondgluemethod-thecyanoacrylateisvaporized,andtheresidualsubstancesonthefingerprint,suchasaminoacidsandglucose,willreactwiththevaporizedglueandmakethefingerprintappear.
(1)Breathobservationmethod:Observethebreathingonthesurfaceofasmoothobject;
(2)Visualobservationmethod:Observewithacertainlightandangle.Withthehelpofbacklighting,fingerprintscanoftenbedisplayed.Thisisbecausefingerprintsareoftenstainedwithdust,whichabsorbslightandappearsdark.
(3)Magnifyingglassobservationmethod:observeunderacertainlightandanglewithamagnifyingglass;
(4)Ultravioletobservationmethod:observewiththecharacteristicsofultravioletrays;
(5)Physicalandchemicalobservationmethod:Forpotentialfingerprintsthataredifficulttoobservewiththenakedeye,physicalandchemicaltreatmentsareneededtofindout.
(6)Cyanidegluemethod:firstprepareasolutionofcyanideglueandether,immersethefilterpaperofasuitablesizeinthesolution,takeitoutanddry,andcontactthefilterpaperwiththefingerprintfor5-60minutes.Thepartofthefingerprintclosetothepapervolatilizesduetotheeffectoftheglue,andthepatternappearsonthesurfaceofthefilterpaper.
Specialfingerprints
Peoplewithspecialdiseasesleave"specialfingerprints"atthescene,whicharemainlyseenincertaindiseasesthatcanchangethepatient'ssweat.Likediabetes,becausethesugarcontentinthesweatofthepatientincreases,ifalotofsweatleavesfingerprints,theremaybeaphenomenonofantsandbeesgatheredinsomenovelsorpuzzles.Also,asmentionedonTVsometimeago,somepeopleusedinferiorporcelainteacupstodrinkteaforalongtime,resultingincopperpoisoning,whichresultedinredsweats.Forpatientslikethis,iftheyleavefingerprints,theywillfindthatthefingerprintsarered.
InArgentina,usingfingerprintevidencetomakeawomanwhokilledhertwochildrenconfessedtoacrime,thisisthefirsttimethatmodernfingerprintinspectiontechnologyhasbeenadoptedbyacourt.
Withthedevelopmentofscienceandtechnology,fingerprintshavenewusesinmedicine.Somedoctorshavediscoveredthatcertaindiseasescanbedetectedbycheckingaperson'sfingerprintsandpalmprints.
Fingerprintshavebecomegoodfriendswithelectroniccomputers.
Manybusinessesalsousethecharacteristicsoffingerprintstodevelopsomehigh-techequipmenttoreflecttheconvenienceandsafetythatfingerprintsbringtolife,suchasfingerprintlocks,fingerprintaccesscontrol,fingerprintattendancemachines,fingerprintcollectionDevice,fingerprintsafe,networkfingerprintlogintechnologyandsoon.Accordingtothesurvey,manyhigh-endsmartcommunitiesinChinaareequippedwithfingerprintlocksandfingerprintaccesscontrol.Theearliestfingerprintsusedintheequipmentisthefingerprintattendancemachine.Inordertopreventthegenerationofcards,thecompany'spersonnelmanagershaveadoptedfingerprintattendancemachinesoneafteranother.Atthesametime,mycountry'sfirstnetworkfingerprintlogintechnologyproviderhaslaunchedabetaversion,whichisexpectedtosolvetheproblemofnetworkaccountsecurity.
Whatnewuseswillsmallfingerprintshaveinthefuture?Anewmazeisbeforeusagain,waitingforustoexploreandseek.
Animalfingerprints
Thestudyfoundthatinadditiontohumans,primatessuchasgorillas,chimpanzees,orangutansalsohaveskinpatternsonthehandsandfeet,andeventhearborealkoalaSkinpatternsarealsofoundonanimalsofitskind(Possumidae).Inaddition,spidermonkeysandcapuchinmonkeyslivinginSouthAmericahaveacurlytail,whichcandexterouslygraspobjects,andtheyalsohaveskinpatternsontheinsideoftheirtails.Itcanbeinferredfromthisthatitisalwayseasiertoformskinlinesontheskinwhereananimalisgrabbinganobject.
Principlesofrecognition
Fingerprintrecognition
Readfingerprintimages,extractfeatures,savedataandcompare.Atthebeginning,theimageofthehumanfingerprintisreadbythefingerprintreadingdevice.Afterthefingerprintimageisobtained,theoriginalimagemustbepreliminarilyprocessedtomakeitclearer.Next,thefingerprintrecognitionsoftwareestablishesthedigitalrepresentationofthefingerprint-characteristicdata,aunidirectionalconversionthatcanconvertfromfingerprinttocharacteristicdatabutnotfromcharacteristicdatatofingerprint,andtwodifferentfingerprintswillnotproducethesameCharacteristicdata.
Somealgorithmscombinethenodeanddirectioninformationtogeneratemoredata,whichindicatestherelationshipbetweeneachnode,andsomealgorithmsalsoprocesstheentirefingerprintimage.Inshort,thesedata,usuallycalledtemplates,aresavedas1Krecords.Nomatterhowtheyarecomposed,thereisstillnostandardfortemplatesandnopublishedabstractalgorithms.Instead,eachmanufacturerdoesitsownthing.Finally,throughthemethodofcomputerfuzzycomparison,thetemplatesofthetwofingerprintsarecompared,andthedegreeofsimilaritybetweenthemiscalculated,andfinallythematchingresultofthetwofingerprintsisobtained.Fingerprintsareactuallymorecomplicated.
Unlikemanualprocessing,manybiometrictechnologycompaniesdonotdirectlystorefingerprintimages.Overtheyears,manydigitalalgorithmshavebeenproducedinvariouscompaniesandtheirresearchinstitutions(therelevantlawsoftheUnitedStatesbelievethatfingerprintimagesarepersonalprivacy,sofingerprintimagescannotbestoreddirectly).
Fingerprintrecognitionalgorithmsultimatelyboildowntofindingandcomparingfingerprintfeaturesonfingerprintimages.ThecharacteristicsoffingerprintsWedefinetwotypesofcharacteristicsoffingerprintstoverifyfingerprints:overallcharacteristicsandlocalcharacteristics.Generalfeaturesrefertothosefeaturesthatcanbedirectlyobservedbythehumaneye,including:basicpatternpatterns,loop,arch,andwhorl.Otherfingerprintpatternsarebasedonthesethreebasicpatterns.Relyingonthepatterntypetodistinguishfingerprintsisfarfromenough.Thisisonlyaroughclassification,butclassificationmakesitmoreconvenienttosearchforfingerprintsinlargedatabases.
PatternArea
(PatternArea)Thepatternareareferstotheareaonthefingerprintthatincludestheoverallcharacteristics,thatis,itcanbedistinguishedfromthepatternareathatthefingerprintbelongstoWhichtype.Somefingerprintrecognitionalgorithmsonlyusethedatainthepatternarea.Aetex'sfingerprintrecognitionalgorithmusestheobtainedcompletefingerprintinsteadofjustthepatternareaforanalysisandrecognition.
CorePoint(CorePoint)Thecorepointislocatedattheprogressivecenterofthefingerprintpattern.Itisusedasareferencepointforfingerprintreadingandfingerprintmatching.
Triangularpoint(Delta)Thetriangularpointislocatedatthefirstbifurcationorbreakpointfromthecorepoint,orwherethetwostripedroadsconverge,isolatedpoints,orturningpoints,Orpointtothesesingularities.Thetriangularpointprovidesthestartingpointforcountingandtrackingfingerprintlines.
TypeLines(TypeLines)Thetypelinesarecrosslinesthatappearwherethelinessurroundingthepatternareabegintobeparallel.Thestylelinesareusuallyshortandinterrupted,butItsouterlinebegantoextendcontinuously.
Numberoffingerprints
(RidgeCount)referstothenumberoffingerprintsinthepatternarea.Whencalculatingthenumberoffingerprints,usuallyconnectthecorepointandthetrianglepointfirst,andthenumberofintersectionsofthislinewiththefingerprintlinecanberegardedasthenumberoffingerprints.LocalfeaturesLocalfeaturesrefertonodesonthefingerprint.Twofingerprintsoftenhavethesameoverallcharacteristics,buttheirlocalfeatures-nodes,cannotbeexactlythesame.Nodes(MinutiaPoints)Fingerprintlinesarenotcontinuous,smoothandstraight,butofteninterrupted,bifurcatedorDiscount.Thesebreakpoints,bifurcationpoints,andturningpointsarecalled"nodes."Itisthesenodesthatprovidefingerprintconfirmationnodecharacteristics
1.Classification-Thereareseveraltypesofnodes,themosttypicalonesareendpointsandbifurcationpoints
A.Endpoints(Ending)-astripedroadendshere.
B.Bifurcation-astriperoadisdividedintotwoormorelineshere.
C.RidgeDivergence-Twoparallellinesseparatehere.
D.DotorIsland(DotorIsland)-aparticularlyshortline,sothatitbecomesapoint.
E.Enclosure-Afterastripedroadisdividedintotwo,itismergedintooneimmediately,andasmallringformedinthiswayiscalledaringpoint.
F.ShortRidge(ShortRidge)-apatternthatisshortatoneendbutdoesnotbecomeapoint.
2.Orientation-thenodecanfaceacertaindirection.
3.Curvature---describethespeedatwhichthegraindirectionchanges.
4.Position(Position)-thepositionofthenodeisdescribedby(x,y)coordinates,whichcanbeabsolute,orrelativetoatrianglepointorafeaturepoint.
Howtoobtain
First,opticalrecognitiontechnology
Theuseofopticaltechnologytocollectfingerprintsistheoldestandmostwidelyusedtechnology.Putyourfingerontheopticallens,andusetheprismtoprojectthefingeronthechargecoupleddevice(CCD)undertheilluminationofthebuilt-inlightsource,andthenformtheridges(thelineswithacertainwidthanddirectioninthefingerprintimage),whichareblackandvalleys.Theline(theconcavepartbetweenthelines)isawhitedigitizedmulti-grayfingerprintimagethatcanbeprocessedbythefingerprintdevicealgorithm.
Second,Temperaturesensorrecognitiontechnology
Itsadvantageisthatfingerprintimagescanbeacquiredwithin0.1s,andthesensorhasthesmallestvolumeandarea,whichisusuallyTheso-calledslidingfingerprintreaderadoptsthistechnology.Thedisadvantageis:subjecttotemperaturelimitations,afteralongtime,thefingerandthechipareatthesametemperature.
3.Semiconductorsiliconsensingtechnology(capacitivetechnology)
SemiconductorcapacitancesensorisbasedonthecapacitancevalueformedbytheridgeandridgeofthefingerprintandthesemiconductorcapacitancesensingparticlesThesizeisdifferenttodeterminewhichpositionisthecrestandwhichpositionistheyu.Itsworkingprocessistopre-chargethecapacitivesensingparticlesoneachpixeltoacertainreferencevoltage.Whenthefingertouchesthefingerprintperformanceofthesemiconductorcapacitor,becausetheridgeisconvexandtheridgeisconcave,accordingtotherelationshipbetweenthecapacitancevalueandthedistance,differentcapacitancevalueswillbeformedattheridgeandtheridge.Thenusethedischargecurrenttodischarge.Becausetheridgeandyuhavedifferentcapacitancevalues,theirdischargespeedsarealsodifferent.Thepixelsunderthecrest(highcapacitance)dischargemoreslowly,whilethepixelsundertheridge(lowcapacitance)dischargefaster.Dependingonthedischargerate,thepositionsoftheridgesandridgescanbedetectedtoformfingerprintimagedata.
Four.Ultrasonictechnology
Theultrasonicfrequencyusedbytheultrasonictechnologyis1×104Hz-1×109Hz,theenergyiscontrolledtotheextentthatitdoesnotdamagethehumanbody(thesameintensityasthemedicaldiagnosis).Ultrasonictechnologyproductscanachievethebestaccuracy.Itrequireslesscleanlinessoffingersandplanes,butitscollectiontimewillbesignificantlylongerthantheaforementionedtwotypesofproducts,anditisexpensive,anditcannotbeusedforlivefingerprintidentification.rare.
Importantconcepts
AsanewITtechnologyfield,fingerprintrecognitiontechnologyhasmanynewconcepts.Understandingtheconceptoffingerprintidentificationtechnologyhelpstounderstandfingerprintidentificationtechnologyaccurately.
Identificationsystem
Afterthedevelopmentofthefingerprintidentificationsystemfrommanualidentificationtomachineidentification,itenterstheautomaticidentificationstage,whichiscalledtheautomaticfingerprintidentificationsystem(AFIS).Atypicalautomaticfingerprintidentificationsystemincludesafront-endsubsystemthatinteractswithpeople-automaticfingerprintacquisitionequipment,aback-endsubsystemthatcompletesfingerprintimageprocessingandfeaturevalueextraction,andadatabasesubsystemforfingerprintlibrarystorage.Whenthebackgroundsubsystemisusedinthefingerprintregistrationprocess,itcanbecalledthefingerprintregistrationsubsystem.Whenitisusedinthefingerprintidentificationprocess,itiscalledthefingerprintidentificationsubsystem.
Registrationmatching
Fingerprintregistrationisalsocalledfingerprintregistration.Thisistheprocessofextractingfingerprintfeaturevaluesfromthefingerprintimage,formingafingerprintfeaturevaluetemplate,combiningitwiththeperson'sidentityinformation,andstoringitinthefingerprintidentificationsystem.Itisequivalenttoreportinganaccountforfingerprints.Therefore,whenregisteringfingerprints,itisnecessarytoensurethecorrectcorrespondencebetweenfingerprintsandidentityinformation.Inparticular,itisveryimportanttopreventimpostersandavoiderrorsintheassociationoffingerprintsandidentityinformationwhenthegovernment,associations,companiesandotherunitsperformfingerprintregistration.Therefore,inthistypeoffingerprintapplication,theprocessoffingerprintregistrationrequirestheparticipationofon-sitesupervisors.Itevencollectsthesupervisor’sfingerprintsintothesystemasacomponentoftheregistrant’sfingerprintfeaturevaluetemplatetoshowtheimportanceofresponsibilitiesandprovideabasisforfollow-upresponsibilityaudits.
Identificationandverification
Identificationandverificationisnotaprobleminthefieldoffingerprintrecognitionalgorithms,butaprobleminthefingerprintrecognitionsystem.Fingerprintrecognitionreferstomatchingfingerprintfeaturevaluesin1:Nmode.Itistheprocessofidentifyingaspecificfingerprintfrommultiplefingerprinttemplates.Theresultis"Yes"or"No."Sometimesinformationabout"whois"isgiven.
Fingerprintverificationreferstomatchingfingerprintfeaturevaluesin1:1mode.Itisaprocessofmatchingthefingerprintfeaturetemplatetobecomparedwithanotherfingerprintfeaturetemplatethatexistsinadvance.Theresultis"isitright".Either1:1modeor1:Nmodecanbeusedinasystem,whichdependsonthecharacteristicsandrequirementsoftheapplicationsystem.Sometimesyoucanalsoconvertthe1:Nmodetothe1:1modetoimprovethesystemsecurityandcomparisonspeedasrequiredbythebusinessmodel.
Advantages:
1.Fingerprintsaredifferentfeaturesofthehumanbody,andtheircomplexityissufficienttoprovidesufficientfeaturesforidentification.
2.Ifyouwanttoincreasereliability,youonlyneedtoregistermorefingerprintsandidentifymorefingers,uptoten,andeachfingerprintisnotthesame.
3.Scanningfingerprintsisfastandveryconvenienttouse.
4.Whenreadingafingerprint,theusermusttouchthefingerandthefingerprintacquisitionheaddirectly.
5.Contactisthemostreliablewaytoreadhumanbiologicalcharacteristics.
6.Thefingerprintcaptureheadcanbemoreminiaturized,andthepricewillbelower.
Disadvantages:
1.Somepeopleorcertaingroupshavefewfingerprintfeaturesandaredifficulttoimage.
2.Inthepast,theuseoffingerprintsincriminalrecordsmadesomepeopleafraidof"recordingfingerprints."
3.Infact,thefingerprintidentificationtechnologycannotstoreanydatacontainingthefingerprintimage,butonlystoretheencryptedfingerprintcharacteristicdataobtainedfromthefingerprint.
4.Everytimeafingerprintisused,theuser'sfingerprintprintwillbeleftonthefingerprintcapturehead,andthesefingerprinttracesmaybeusedtocopyfingerprints.
FRRandFAR
FRR(FalseRejectionRate)andFAR(FalseAcceptanceRate)arethetwomainparametersusedtoevaluatetheperformanceoffingerprintrecognitionalgorithms.FRRandFARaresometimesusedtoevaluatetheperformanceofafingerprintrecognitionsystem,butthisisnotappropriate.Inadditiontotheperformanceofthefingerprintidentificationsystembeingaffectedbythefingerprintalgorithm,theimpactoftheperformanceofthefingerprintcollectiondeviceonFRRandFARcannotbeignored.
FRRiscommonlycalledfalserejectionrate,andthestandardnameisFNMR(FalseNon-MatchRate,mismatchrate).Itcanbeeasilyunderstoodastheprobabilityof"takingfingerprintsthatshouldbesuccessfullymatchedwitheachotherasfingerprintsthatcannotbematched".Theperformancemeasurementofthefingerprintalgorithmismeasuredundertheconditionofagivenfingerprintlibrary.ThefingerprintlibraryusedformeasurementisgenerallygivenbytheorganizerofFVC(InternationalFingerprintIdentificationAlgorithmCompetition).WhenFVCperformsthefingerprintrecognitionalgorithmperformancetest,thereisnoexternalfingerprintinput,anditusesastandardfingerprintimagelibrarytotest.Therefore,FNMRisatestvalueobtainedwithoutconnectingafingerprintcollectiondevice.Theotherparametersinthissectionarealsoderivedunderthispremise.
Assumingthatthereare100fingerswithdifferentIDsinthefingerprintlibrary,andeachfingerisregisteredwith3fingerprints,thereareatotalof300fingerprintsinthefingerprintlibrary.AssumingthatP1representstheIDoffinger1,thefingerprintregisteredthreetimesisP1-F1,P1-F2,P1-F3.FNMRreferstocomparing3fingerprintsofthesamefingerinthefingerprintdatabasepairbypair,namelyP1-F1andP1-F2matches,P1-F1andP1-F3Match,P1-F2matchesP1-F3,P1-F2matchesP1-F1,P1-F3matchesP1-F1,P1-F3matchesP1-F2,Thereare6matchingmethods.Make6kindsofmatchesforall100fingersinside,atotalof6×100=600matches.Intheory,600matchescanbematchedcorrectly,andthesuccessrateofmatchingis100%.Infact,becausethethreefingerprintimagesofthesamefingercannotbeexactlythesame,thereisaproblemofmatchingsimilarity.Supposewesetthesimilarityofsuccessfulmatchingto>90%,thatis,whenthesimilarityisgreaterthan90%,itmeansthatthematchingissuccessful.Thenwefindouthowmanytimesthesimilarityisabove90%fromthe600matches.Thisvaluerepresentsthenumberofsuccessfulmatches,whichisassumedtobe570.Therestofthe600timesindicatethenumberoftimesthatthereisnosuccessfulmatch,whichis600-570=30times.Thematchingfailurerateis30/600=5%.
Forthefingerprintrecognitionalgorithm,whenthefingerprintdatabaseisconfirmed,thematchingfailurerateFNMRiscertain.Whenthefingerprintlibrarychanges,itsFNMRwillalsochange.Therefore,internationally,thefingerprintlibrarypublishedbyFVCisusedasaunifiedtestlibrary,andtheFNMRresultstestedinthetestlibraryareusedasastandardreferenceformeasuringtheperformanceoffingerprintalgorithms.
FARisgenerallycalledthefalserecognitionrate,anditsstandardnameisFMR(FalseMatchRate).FMRisthemostimportantparameterusedtoevaluatetheperformanceoffingerprintrecognitionalgorithms.Itcanbeeasilyunderstoodastheprobabilityof"takingfingerprintsthatshouldnotbematchedasmatchedfingerprints".
Alsotakethefingerprintlibraryinthepreviousparagraphasanexample.Matcheachfingerprintinthelibrarywithallotherfingerprintsexceptyourself.Thetotalnumberofmatchesis300×(300-1)=89700times.Intheory,thenumberofsuccessfulmatchesis6×100=600times,andthenumberoffailedmatchesshouldbe89700-600=89100times.Supposethatduetotheperformanceofthefingerprintalgorithm,thematchthatshouldhavefailedisjudgedasasuccessfulmatch,ifitisassumedthatthenumberofsucherrorsis100times.ThefalseacceptancerateFARis100/89100=0.11%.Thenumberofmatchingfailuresvariesduetothestrictnessofthesimilarconditions.Whenthefilterconditionthatmatchessuccessfully,thatis,thethresholdvalueincreases,theFARwilldecrease.
FARisalsorelatedtofingerprintlibrary.IntheFVCcompetition,thereare4fingerprintlibrariesfortesting,andtheaverageistaken.Oneofthefingerprintlibrariesisartificiallygeneratedtoeliminatetheeffectofdifferentfingerprintimagequalitycausedbydifferentcollectiondevicesontheperformanceofthealgorithm.
Inthesamefingerprintlibrary,forthesamealgorithm,athresholdneedstobesetasacriterionforjudgingsimilarity.Whenthesimilarityisgreaterthanthisthreshold,itmeansthematchissuccessful,otherwiseitmeansthematchfails.FNMRincreaseswiththeincreaseofthethreshold,thatis,thehigherthethresholdforjudgingsimilarity,thegreatertheprobabilitythatatruefingerprintisjudgedtobefalse.Conversely,theFMRdecreaseswiththeincreaseofthethreshold,thatis,asthethresholdforjudgingthesimilarityishigher,theprobabilityofjudgingthefakefingerprintastruewillbesmaller.FARisinverselyproportionaltoFRR.AccordingtothetestresultsoftheFVCcompetitionin2004,whentheFMRis1/1000,theFNMRisabout5/100.Thatis,inthefingerprintlibraryof100fingers,if1000matchesareperformed,amatchingerrormayoccur,thatis,anacknowledgment.After100matchesareperformed,theremaybe5matchfailures,thatis,theyarenotrecognized.
EER
EER(EqualErrorRate)meansequalerrorrate.Thisparameterisgenerallynotusedinordinaryoccasions.EERismainlyusedtoevaluatetheoverallperformanceoffingerprintalgorithms.Thatis,thetwoparametersofFARandFRRareunifiedintooneparametertomeasuretheoverallperformanceofthefingerprintalgorithm.FARandFRRaretwoparametersofthesamealgorithmsystem.Puttheminthesamecoordinate.FARdecreaseswiththeincreaseofthethreshold,andFRRincreaseswiththeincreaseofthethreshold.Sotheymusthaveintersections.ThispointisthepointatwhichFARandFRRareequivalentunderacertainthreshold.Itiscustomarytousethisvaluetomeasuretheoverallperformanceofthealgorithm.Forabetterfingerprintalgorithm,itishopedthatbothFARandFRRshouldbeassmallaspossibleunderthesamethreshold.
TranslateboththeFARandFRRcurvesdownward.Atthesametime,theintersectionpointERRalsoshiftsdownward.ThesmallertheEERvalue,thehighertheoverallperformanceofthealgorithm.
BecausethecorrespondingthresholdswhenFRRandFARintersectareverysmall,thatistosay,thesimilaritythresholdatthistimeislessthan30%.Thethresholdinactualuseisatleast80%,sotheEERvalueisnotusedinpopularoccasionstodescribetheperformanceofthefingerprintalgorithm,butonlyusedincompetitionrankings.
FRRisactuallyanimportantindicatorofsystemeaseofuse.SinceFRRandFARarecontradictorytoeachother,thismakesitnecessarytoweigheaseofuseandsafetyinthedesignofapplicationsystems.Aneffectivemethodistocomparetwoormorefingerprints,therebygreatlyimprovingsystemsecuritywithoutlosingeaseofuse.
Disapprovalrate
Thedisapprovalrateisgenerallyseldomused.Infingerprintrecognitionterms,itisarelativelyvagueterm.IntheWorldFingerprintAlgorithmCompetition,thereisaparametercalledtherejectionrate,sometimescalledtherejectionrate,whichisusedtomeasurethedegreeofpickinessofthefingerprintrecognitionalgorithmonthequalityofthefingerprintimage,usingREJENROLL.Express.Inagivennumberoffingerprints,suchas100fingerprintimages,fingerprintsthatcanbesuccessfullyregisteredorcalledfilecreation,ifitis99,thenREJENROLL=1%.ForthestandardfingerprintlibrarygivenbytheFVCcompetition,mostfingerprintalgorithmscanbesuccessfullyfiled,thatis,REJENROLLis0.00%.
Onanotheroccasion,therejectionrateisusuallyinterpretedastheprobabilitythatthefingerprintidentificationsystem(includingfingerprintcollectionequipment)willnotacceptfingerprintregistration.Inthiscase,thefactorsthatrejectregistrationaremoreaffectedbytheimagingcapabilitiesofthefingerprintcollectiondevice,inadditiontothealgorithmitself.Thebetterthequalityofthefingerprintimageoutputbythefingerprintcollectiondevice,thelowertherejectionrateofthefingerprintidentificationsystem,andthelowerthequalityofthefingerprintimageoutputbythefingerprintcollectiondevice,thehighertherejectionrate.
Time
Registrationtimeisanotherindicatorusedtomeasuretheperformanceoffingerprintalgorithm.Itreferstothetimefromtheinputofthefingerprintimagetothesuccessfulestablishmentofthefingerprintfile(successfulregistration).AccordingtotheresultsoftheFVCcompetition,thegeneralfingerprintalgorithmregistrationtimeiswithin0.5seconds,whichisalsooneofthequalificationsproposedbyFVCtoparticipateintheLIGHTgroupcompetitionalgorithm.
Thematchingtimeissometimescalledthecomparisonspeed,whichisusedtoindicatethetimerequiredforthefingerprintrecognitionalgorithmtocompleteamatch.Itisthetimefromthefingerprintimageinputtotheoutputofthematchingresult.Thematchingtimeofmostalgorithmsparticipatinginthealgorithmcompetitioniswithin0.3seconds.ThisparameterandtheminimumregistrationtimeconstitutetheentryconditionsfortheLIGHTgroup.
Becausethesetimesareallaffectedbythequalityofthefingerprintimagetobetested,theaveragevalueofmultiplefingerprintlibrariesisgenerallytaken,andtheaverageregistrationtimeandaveragematchingtimearegenerallyusedasthebasisformeasurement.
Technologyintroduction
Theresearchfrom"fingerprint"to"fingerprinttechnique"hasgonethroughalongprocess.Aftertheformationoffingerprinttechnology,ithasundergonethedevelopmentandtransformationfrommanualidentificationtechnologytoautomaticidentificationtechnology.Withthedevelopmentofcomputerimageprocessingtechnologyandinformationtechnology,fingerprintrecognitiontechnologyhasgraduallyenteredthefieldofITtechnology,combinedwithmanycomputerinformationsystems,andwidelyused.
Expertsreminded
ProfessorsfromtheNationalInstituteofInformaticsinJapanremindednetizensthatifyouposeaVsignwhentakingpictures,yourfingerprintsarelikelytobestolen.Inadditiontofingerprints,facialandirisrecognitionhasalsobeenappliedtomobilephoneauthentication.Someadministrativeagenciesandenterprisesarealsousingthisinformationforattendancemanagement.Previously,inordertoobtainpersonalbiometricinformation,itwasnecessarytoapproachthepersonforshooting.However,recentlysomebiologicalinformationhasbeencirculatedontheInternet,whichgreatlyreducesthethresholdforcriminals.