HAL Id: hal-01998605
https://hal.archives-ouvertes.fr/hal-01998605
Submitted on 29 Jan 2019
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of
sci-entific research documents, whether they are
pub-lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Laser-induced fluorescence study of medieval frescoes by
Giusto de’ Menabuoi
Roberta Fantoni, Luisa Caneve, Francesco Colao, Luca Fiorani, Antonio
Palucci, Ramiro Dell’Erba, Vasco Fassina
To cite this version:
Italian National Agency for New Technologies, Energy and
Sustainable Economic Development
http://www.enea.it/en
http://robotica.casaccia.enea.it/index.php?lang=en
This paper is a pre-print. The final paper is available on:
Author's personal copy
JournalofCulturalHeritage14S(2013)S59–S65
Available
online
at
www.sciencedirect.com
Laser-induced
fluorescence
study
of
medieval
frescoes
by
Giusto
de’
Menabuoi
Roberta
Fantoni
a,∗,
Luisa
Caneve
a,
Francesco
Colao
a,
Luca
Fiorani
a,
Antonio
Palucci
a,
Ramiro
Dell’Erba
b,
Vasco
Fassina
caENEATechnicalUnitforthedevelopmentofapplicationsofradiations,CRFrascati,V.E.Fermi45,00044Frascati,Italy
bENEATechnicalUnittechnologiesforenergyandindustry–RoboticsLaboratory,CRCasaccia,ViaAnguillarese301,00123Roma,Italy cSoprintendenzaaiBeniStoriciArtisticiedEtnoantropologiciperleprovincediVenezia,Belluno,PadovaeTreviso,RioMarin,30124Venezia,Italy
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:Received24October2012 Accepted31October2012 Availableonline20February2013 Keywords:
Laser-inducedfluorescence Frescoes
PCanalysis
Spectralanglemapper Giustode’Menabuoi
a
b
s
t
r
a
c
t
Laser-inducedfluorescence(LIF)isapowerfulremoteandnon-invasiveanalysistechniquethathasbeen successfullyappliedtothereal-timediagnosisofhistoricalartworks.Hyperspectralimagescollectionon fresco’sandtheirfalsecoloursprocessingallowedtorevealfeaturesinvisibletothenakedeyeandto obtainspecificinformationonpigmentscompositionandconsolidantsutilization,thelatteralsorelated toformerrestorations.ThisreportpresentstheresultsobtainedbyENEALIFscanningsystemduringa fieldcampaignconductedinJune2010onfresco’sbyGiustode’MenabuoiinthePaduaBaptistery.The datacollectedbyLidArtallowedthedetectionofParaloidB72andMovilith/PrimalAC33,guidingthe restorersintheirconservationactions.
©2013ElsevierMassonSAS.Allrightsreserved.
1. Introduction
Laser-inducedfluorescence(LIF)isapowerfulremoteanalysis toolthathasbeensuccessfullyappliedtothereal-timediagnosisof historicalartworks,allowingtheobservationoffeaturesinvisible tothenakedeye,aspigmentcomposition[1],biologicalattack[2]
andtraceofformerrestorations.
InatypicalLIFinstrument,anultraviolet(UV)laserbeam irra-diatesasampleandanopticalsystemmeasuresthefluorescence spectrum that contains informationonthe target composition. Nevertheless,alternative spectroscopic techniques,suchas XRF andmicro-Ramanaresuitabletoobtainanalogousinsitu informa-tion[3].LIFisuptonowtheonlyoneforwhichremoteapplication hasbeendemonstrated up toten of meters.LIF is indeed fast, remote[4],non-invasive,sensitiveandselective.Theseadvantages have encouragedits application in the real-time monitoringof historicalfrescoes,mosaics,paintingsandstones.Formerstudies showeditshighpotentialasadiagnostictoolforculturalheritage eitheroperatinginspectral[1,2]ortime[5]resolvedmode.
TheTechnicalUnitforthedevelopmentofapplicationsof radi-ations,partoftheItalianNationalAgencyforNewTechnologies, EnergyandSustainableEconomicDevelopment(ENEA)hasworked inthisfieldsince2003andhasdevelopedvariousprototypes.The systemheredescribed(LidArt)isalidarfluorosensor(opticalradar basedonLIF)and itsfirstversion waspatented in2007[6]. At
∗ Correspondingauthor.
E-mailaddress:fantoni@frascati.enea.it(R.Fantoni).
present,itisevenmorecompact(containedinacylinderofradius 29cmandheight18cm),hyperspectralandtimeresolved(the flu-orescencespectrumismeasuredwithawavelengthresolutionof few nm and a timeresolution of few ns).Moreover,it is light, robust,transportable,user-friendlyandcost-effective.Eventually, itsscanning(theartworksurfaceisprobedremotelyaimingthe laserbeam)isbasedonaninnovativeapproach[7]thatreduces theacquisitiontimefromhourstominutes.
Thanks tosophisticated dataprocessing techniquesas false-colorimaging,principalcomponentanalysis(PCA)onspectraand spectralanglemapping(SAM)onimages,LidArthasdetected char-acteristicsinvisibletothenakedeye,aspigmentcompositions(e.g. titaniumwhitevs.zincwhite[8]),pigmentdiffusions(limeand casein)[9],biologicalattacks(algaeandfungi)[10],consolidants (usuallyresins)[11],deteriorations,depigmentations, retouches andvarnishes.
In particular,restorers needa precise knowledge of surface materialsonceengagedinremovingtracesofformerrestorations. LIFimages,collectedinthePaduaBaptisteryandprocessedinorder toretrieveinformationonconstituentmaterials,areherepresented anddiscussed.ThefieldcampaignwascarriedoninJune2010prior totheplanningofsuccessiverestorations.Tothisaim,the investiga-tionwasfocusedontotherecognitionofconsolidantsandbinders, withthesupportofanavailabledatabaseandthedevelopmentof statisticaltoolsformainspectralbandsidentification.
TheconstructionofPaduaBaptistery,locatedontherightside ofthecathedral,startedin12thcenturyandwasaccomplishedin 1281.Itsfresco’sareconsideredGiustode’Menabuoi’smasterpiece. Inthepresentstudy,carriedonwithinacontractsignedbyENEA
1296-2074/$–seefrontmatter©2013ElsevierMassonSAS.Allrightsreserved.
S60 R.Fantonietal./JournalofCulturalHeritage14S(2013)S59–S65
andSoprintendenzaaiBeniStoriciArtisticiedEtnoantropologiciper leprovincediVenezia,Belluno,PadovaeTreviso,selectedareasof theParadisefrescoonthecentraldomeandoftheGenesisonthe tambourhavebeeninvestigatedbymeansoftheENEALIFscanning system.
Aftera shortdescriptionofthelastLidArt prototype,results obtainedduringthefieldcampaignconductedfromJune7to11 arepresentedanddiscussed.Someconclusionsonthe appropri-atenessofthedevelopedtechnologicaltooltotherequiredremote investigationonapaintedCHsurface,aregivenattheend.
2. Experimental
2.1. Theset-up
The LIF scanningsystem developed and patented [6] atthe ENEAlaboratoryinFrascatiiscapabletocollecthyperspectral flu-orescenceimagesscanninglargeareasforapplicationstocultural heritagesurfaces(e.g.frescoes,decoratedfacadesetc.).Anew com-pactset-uphasbeenbuilt,andsuccessivelypatented[7],aimed toincrease theperformancesin termsofspaceresolution, time resolvedcapabilitiesand dataacquisitionspeed.Major achieve-mentshavebeenreachedbyacriticalreviewoftheopticaldesign andconsequentlyofthedetectorutilized:thepointfocalization modehasbeenchangedintoalinefocalizationbyusingaquartz cylindricallensandaimagingspectrograph(Jobin-YvonCP240); thelineararraydetector,responsibleforthemultichannelspectral resolution,hasbeenreplacedwithasquareICCDsensor(ANDOR iStarDH734,pixelsize13m),mountedbehindaslitparallelto thelaserlinefootprintduringthescanning.
This arrangementis characterizedby havingthespatial and spectral information on two mutually orthogonal directions imagedonthedetector,withsubmillimetricspatialresolutionand aspectralresolutionbetterthan2nm.Additionally,itispossibleto implementtimeresolvedmeasurementsonthenanosecondscale bygatingthespectraldetector.
Theoverallcurrentsystemperformancesarehorizontal resolu-tion640pixel,0.1mradangularresolution,minimumacquisition timeperline200ms,fieldofview(FOV)aperture5.7◦ (correspond-ingtoascannedlineof2.5mat25mdistance).Withthelatter optics,animageof1.5×5m2isscannedinlessthan2minutesat
25m.
The compact arrangement of the experimental apparatus is showninFig.1,togetherwiththeadoptedoptical scheme.The Nd:YAGcompactlasersourcecouldbeoperatedattwodifferent
wavelengthsintheUV:355nmor266nm,1mJofoutputenergy wereproducedinshortpulses(10ns)at20Hz.Specialcarehasbeen paidtocheckonreferencesamplesinlaboratorythattheutilized laserfluencewaslowenoughtoavoidanyphoto-damageinduced bythelasertothepaintedsurface(energydensity<7J/cm2)as
testedonfrescospecimenstothepurposerealized.Inparticular measurementswerecarrieddefocussingthelaserbeam.The com-binationof(low)laserrepetitionratewith(high)scanningspeed allowedtoavoidtheoccurrenceofanyphotochemicalprocessat thesurface(e.g.pigmentoxidationorconsolidantdecomposition) relatedwithcontinuousirradiationofthesamepoints.
2.2. Operatingmodes
Thelinescanningsystemcanoperateintwooperatingmodes byslightlychangingtheactivecomponentsintheentiredevice: i.e.reflectanceandfluorescencemeasurements.Thesemodesshare indeedmostoftheoptoelectronicequipment,namelythe collec-tionopticsandthespectraldetector,whilediffering,ononehand, inthewaythesampleisexcitedand,ontheotherhand,intheway thespectraldetectorisoperated.
Reflectancemeasurementsareperformedinascaninwhichthe laserisswitchedoffwhilethesampleisexposedtothelightemitted byaNISTtraceablelamp.Thecollectionopticsarefocussedontothe scannedsurfaceandtheacquisitioncameraisrunningbylocking itsinternalsynchronismtothepowerline(toavoidimage flicker-ing).Theresultgivesthereflectancespectrumforeachpixelofthe scannedarea,fromwhichtheCIE/labcoordinatescanbecomputed oncethesystemiscalibratedagainstareferencesurface.
Theacquisitionoffluorescencespectrainducedbylaseris car-riedonduringascanwiththelaserswitchedon.Thesystemuses aproceduretorecordspectraalsoinfulldaylight:inthiscase, a proprietarymethodis usedtodiscriminatebetweenthelight inducedfromthelaser(LIFsignal)withrespecttothelight diff-usedbytheareaunderstudyandbytheenvironment.Theresult givesthefluorescencespectrumforeachpixelofthescannedarea; thenasuccessivedataanalysisoftheacquiredspectrumprovides detailedinformationontheareaunderstudy.
LIFfluorescencemeasurementsrequiretheadoptionofa pecu-liarmeasurement protocolinordertoachievea gooddetecting efficiencyandhighspatialresolutionevenintheverybroadspectral rangeofradiationemitteduponlaserexcitationat266nm.Infact, becauseoftheuseofrefractivecollectingoptics,thefocalplanesfor ultraviolet(UV)andvisible(VIS)radiationoccurindifferent posi-tion,causingdefocuseffectsanddegradingthespatialresolution
Author's personal copy
R.Fantonietal./JournalofCulturalHeritage14S(2013)S59–S65 S61
ineitheroftheregions.Toovercometheproblem,successive sep-arateacquisitionrunsarescheduledforeveryscan,eachofthem optimizedforaspecificwavelengthrange.
Actually,threeconsecutiveacquisitionrunsarescheduled:in thefirstthecollectingopticsisfocusedontheUVregionfrom250 to450nm;thesuccessivetwoscansarefocusedonthespectral regionfrom400to750nmtoacquirerespectivelythefluorescence inducedinthevisibleandthereflectanceimage.
2.3. Analysisofspectralmeasurements
Tospeeduptheprocessingofacquiredimages,themost rele-vantspectralfeaturesareidentifiedbyPrincipalComponent(PC) analysis.Althoughitis commonlyadmittedthatthePCsdonot possessanydirectphysicalmeaning,theycanneverthelessbe rep-resentedasspectrasuitabletobedescribedintermsofbands.In somecases,agivenPCcorrespondstowelldefinedspectroscopic bandswithassociatedpeaks,whileinothercasesmore compli-catedtrendsandshapesareobserved:mostfrequentisthecaseof abandsetagainstanother.FewofthePCsareusuallyretainedfor subsequentanalysis:typicallyfivetoeightcomponentsareenough todescribetheentirespectraldataset.Itisalsoworthnoticingthe possibilitytobuildsuitablelinearcombinationsofthecomputed PCstohaveafaithfulrepresentationofeachpixelspectra, even-tuallyusedforthecomputationofstandardCIE/labcolorimetric measurements.
Inthepresentpaper,thePCanalysisisdevoted tothe iden-tificationofprominentspectralfeatures,thusrelievingfromthe lengthytimeconsumingexaminationofeachacquiredspectrum
[12].Thisprocedureisadvantageousbecauseitisfastandcanrunin asemi-automaticmode.However,ithastheinconveniencetogivea globalanalysis,possiblyignoringthoselocalpeculiaritieswhichdo notpossessenoughstatisticalsignificance.Toovercomethis draw-backlocalPCanalysisarealsoperformedondifferentportionof thescannedareas,thentheresultsareanalyzedseparately.Once identified,spectralbandsaresoughtforintheacquiredLIFspectra, completingthedataanalysis.
Adifferentmethod,usedintheanalysisofspectralimages, con-cernstheidentificationofregionshavingaspecificspectralcontent. Typicalisthecaseofidentificationofagivenpigmentinanimage: suchtaskisaccomplishedeitherbyabandanalysis,orbyusing projectionalgorithmslikeSpectralAngleMapper(SAM)or Spec-tralCorrelation Mapper(SCM)incombinationwithanavailable database[13].Althoughthemapperalgorithmsperformwellwith alowcomputationalcost,theirperformancesaregenerallylower withrespecttothebandanalysisprocedures.
3. Results:LIFreferencespectrafromfrescomaterials
Beforepresentingimagesremotelycollectedonfrescoesatthe Baptistery,inordertoproceedwithaneffectivedataanalysis,it isworthwhiletoreportthereferencespectrawhichcanbe uti-lizedforSAMprojectionsinordertoidentifyspecificcontribution fordifferentchemicalsatthefrescosurface.Tothisaimlaboratory sampleswerecharacterizedaspreparedaccordingtotraditional receiptsconcerningbothmaterialsandmethods.Afinallayerof consolidanthasbeenaddedontopofeachsamplesurface.Based onformerinformationofrestorationscarriedoninPadua Baptis-teryinthefollowing,wedescribeonlyapartoftheavailabledata base,includingacrylicresins,bindersandpigments,whicharemost likelyrelevanttothefrescoesuponinvestigation.Notethat nor-malizedspectra,gatherafterdifferentsamplepreparationresulted alwaysverywellcomparable,thussupportingthepeculiar spec-tralfeatureassociatedtoeachcomponentuponmonochromatic UVexcitation. 300 400 500 600 700 800 900 0.0 0.2 0.4 0.6 0.8 1.0 Calaton Mowilith 50 Primal AC33 Paraloid B72 Vinavil background
Fluorescence Intensity (a.u.)
wavelength(nm)
Fig.2.Fluorescencespectraofacrylicresinsonlaboratoryfrescossamplesexcited at266nm.
3.1. LIFreferencespectraofacrylicresins
Sinceabout40yearsacrylicresinswerecommonlyutilizedas consolidantinrestorationofpaintedsurfaces,includingfrescoes. LIFspectroscopyhasbeenutilizedforthecharacterizationoffive differentresins:paraloidB72,calaton,primalAC33,mowilith50 andvinavil.Alaboratorysamplesetwaspreparedduringa coop-erationLjubljanaRestorationCentre.Eachsamplewasformedbya plastersubstrate,apigmentlayerandcoveredbyanacrylicresinat differentthickness.Referencespectrawerecollectedupon excita-tionat355nmand266nm,theformershowedabroadblue-green bandcharacteristicfortheentirefamily,whilethespectracollected fromthelatterareshowninFig.2,theirmainfeaturesare summa-rizedinTable1.
Thespectracollectedafterexcitationat266nmaresuitableto resinsclassification,sincetheyshowspecificemissionbandsfor eachdifferentsample[11];namelyonlycalatonpresentsa char-acteristicblue-greenbroadvisibleemission,whileallothersare strongUVemitters.Amongthemvinavilcanberecognizedbyits slightlybroaderUVbandaccompaniedbyaweakerwelldefined secondpeakat510nm.Finally,thesubstrateemissionaloneis com-pletelydifferent,notshowinganysignatureoverlappingwiththe examinedconsolidants.
3.2. LIFreferencespectraofbinders
Since historical times, organic binders are utilized to keep togetherpigmentspowders.Eightdifferentcompounds,listedin
Table2,wereinterrogatedbyLIFattwodifferentexcitation wave-lengths. Samples wererealized at theChemistry Dep.of Rome UniversitySapienza,byplacingdirectlythebinderontopofanon fluorescentsubstrate.
Referencespectra,collectedat2mdistanceuponlaser excita-tionat355nm,areshowninFig.3,respectivebandcentrepositions
Table1
Bandcentresoffluorescencespectraofacrylicresinsonlaboratoryfrescossamples excitedat266nm.
Resin Bandcentre(nm)
S62 R.Fantonietal./JournalofCulturalHeritage14S(2013)S59–S65
Table2
Pigmentspeakemissionexcitedat266nm.
Pigment Bandcentres(nm)
(exc)=255nm Lineoil 460 Leadwhite 460 Calcite 450,580 Leadsulfate 450 Nopigment(plaster) 460 400 450 500 550 600 650 700 750 800 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 x 104 Wavelength [nm]
LIF intensity [A.U.]
BgndC3 RabbitGlueC2 EggWhiteC1 EggYolk B3 LineseedOil B1 Casein3 A1
Fig.3.Fluorescencespectraofbinderonaninertsubstrateuponexcitationat 355nm.
aregiveninTable3.Dataobtaineduponexcitationat266nmare
notshownsince,apartfromrabbitgluepeakedat310nm,all spec-traappearverysimilarwithamaximumaround340nmandatail alongthevisible.Wemaythusconcludethat,onceinformationon bindersaresearchedfor,itisworthwhiletoutilizethelongerUV wavelength(355nm)todetectspectralfeatureswhichmightallow fortheiridentification.
4. Results:thecampaignonPaduaBaptisteryfrescoes
4.1. Thescannedareas
IntheBaptistery,differentportionsofthefrescoeshavebeen scannedwithinthedomeandonthetambour.Siximageswere collectedonthedomevaultand eight imagesonthetambour. Datawereacquiredat15mdistancefromthedomeataslightly smallerdistancefromthetambour.Imagescollectedareshaped asrectangularstripesabout1.5mbroadandtypicallyfrom8mto 12mlong.
Table3
Binderspeakemissionexcitedat355nm.
Binder Bandcentres(nm)
(exc)=355nm Casein 450 Lineoil 520 Eggwhite 520 Rabbitglue 520 Redwhite 540 Background(silica) 520
Fig.4.SectionsofthedomevaultscannedbyLIF,greylevelscorrespondto differ-entemissionintensity(blackrepresentthehighestintensity);(A)and(B)bandat 293nm;(C)and(D)bandat370nm.
Data wereacquiredoperating theset-upboth in reflectance
andfluorescenceacquisitionmode,thelatteruponexcitationat
266nm.Datawereprocessedbymeansofanautomaticstatistical
analysis(PCA)aimedtorecognizethepresenceofprominentbands.
Oncemajorfluorescencebandshavebeenidentified,their
assign-menthasbeencheckedagainsttheavailableconsolidantdatabase
bymeansofSAMprojections,andtherelevantdistributiononthe
imagehasbeenobtained.
The band analysis, upon which the monochromatic image
reconstructionreportedinFigs.4and5isbased,derivesdirectly
Author's personal copy
R.Fantonietal./JournalofCulturalHeritage14S(2013)S59–S65 S63
Fig.5.(A)FalsecolourRGBreconstructionofasectionofthedomevaultbasedonbandsat293nm,370nmand550nm;(B)averagespectralcontentinselectedspectral regions,whicharemarkedontheleftimage.Notetheintensebandappearsat375nmtentativelyassignedtovinylconsolidants.
Afurtherqualitative resultsofLIF scanningcan beobtained combiningthreesignificantbandsinasinglefalsecolourimages, an example of this elaboration is presented in Fig. 5 (for the monochromaticimageshowninFig.4D).Theinset(A)showthe RGBreconstructionbasedonbandpeakedat550nm,370nmand 293nm;theinset(B)showstheaveragespectruminthe respec-tivecircledareaontheleft,significantemissionsat375nmand at293nm(confirmingPCAresults) areobserved, theirpossible assignmenttodifferentconsolidantsbeingabovediscussed.
TheSAM application gives a preciselocalization of different materials, however this is not true when overlapping contrib-utionsarefoundwithcomparableintensities.Anexampleofthis unfavourablecaseisshowninFig.6:theleftinset(A)showsthe similaritymapreferredtoparaloidB72fluorescencespectrum, cov-eringabout15%oftheexaminedsurface,howeverananalogous percentageisobtainedwhenconsideringthesubstratumspectrum in theright inset(B). The remarkablespacecorrelation among theretrievedmapisanadditionalindicationofthestrong over-lapbetweentheconsideredspectralcontributionwhichcannotbe effectivelydisentangledbytheusedSAMalgorithm.
As a last example of consolidant identification, a vinyl compound,whosefluorescencespectrumisavailableaspure com-pound(Fig.2),hasbeenconsidered.Sincethecompound could notbedepositedonaplastersimilartotheoneutilizedinPadua Baptisteryanditwasnecessarytohaveagoodspectralreference toretrieveareliablespacedistribution,atwo-stepanalysiswas carriedon.First,thesimilarity mapwasproduced withrespect to the purecompound, small areas (about 1%) unambiguously characterizedby itspresence were identified (Fig. 7A).Second, the spectral signature of these areas was assumed as internal
referenceofthecompoundontheusedplasterandthefinal distri-butionwasobtained(Fig.7B),coveringabout11%oftheexamined surface.Althoughthisindirectproceduremightgiveriseto arte-facts,incaseofinitialincorrectassumptionornoisyspectra,and theconfirmationoftheassignmentshouldbeachievedbymeans of complementary techniqueswhenever possible,theproposed methodisa validtoolforanimmediatenon-invasive screening andalsotothemosteffectivelocalizationofsuccessivesampling areas.
4.2. Reflectancemeasurementsonpigments
AnRGBimageobtainedfromreflectancemeasurementsonthe vaultisshown inFig.8A.Spectral featureshavebeenidentified byPCAThereconstructedreflectanceimagerevealsaninteresting feature under Christ’srighthand: a redarea withhighly satu-rated color, whose spectral analysisshows an alterationin the ratiobetweentheredandbluemaxima(at622nmand477nm, respectively).
S64 R.Fantonietal./JournalofCulturalHeritage14S(2013)S59–S65
Fig.6. SAMmapsassociatedwith:paraloidB72(A)andplaster(B).
Fig.7. SAMmapsassociatedto:vinylfromourdatabase(A)andavinylwithan internalreference(B).
not clear, since differentcauses might contribute to its occur-rence:retouches withadifferentredpigment,humidityonthe wall, differentbinders or added varnishes withdiffusive inho-mogeneous reflectance at the surface. Once again, additional complementarymeasurementswouldbenecessaryforadefinite answer, however, the present measurements indicate where samplesshouldbecollectedandhowmuchtheanomalyextends onthepaintedsurface.
Author's personal copy
R.Fantonietal./JournalofCulturalHeritage14S(2013)S59–S65 S65
5. Conclusions
Laboratorymeasurements,aimedtotherealizationofaLIFdata baseforremoteinvestigationoffrescoes,incombinationwiththe developedimageanalysismethodology,havebeenutilizedto dis-cussresultscollectedduringthecampaignofcharacterizationof PaduaBaptisteryfrescoes,asanexampleofmonumentalpainted surface.Presentresultsallowtosummarizemajorachievementsof theproposedinnovativetechnique,asfollows:
• theLIF instrument developed is suitable tonon-invasive and remotesurfacediagnostics;namelyitpermitstoscanlargeareas withmedium/highresolutionandtoreleasedigitalimageswhose false colors canbe associatedwith detectedspectral features observeduponUVexcitationat266nm(e.g.presenceofspecific consolidant/binder);
• theidentificationoftheconsolidant/binder,ispossibleprovided that itsLIF spectrum is present in the availabledata baseas detectedona similarcomplexmatrix;alternatively the pres-enceofa classof consolidants(e.g.acrylic,vinyl, etc.)can be assumedanditsdistributioncanbemeasuredontheentireimage byusingasreferenceandinternalstandardwhosespectrumhas beenobtainedinanareaoftheimagewhereitdominates; • theadditionalavailabilityofreflectanceimages,measuredupon
externalilluminationorutilizinganadditionalvisiblelampwhile keepingthelaserswitchedoff,bymeansofthesamescanning system,givesinformationonpigmentsdistributionand degra-dation.
Thelimitationsofthediagnostictechniquerelymostlyonthe natureoftheCHtargetconsidered,duetoitscomplexitywitha multilayeredstructureoftencontainingmaterialswithsmall chem-icaldifferences.Majordrawbacksforfrescoesarereportedinthe followinglist:
• difficultiestobuilta reliabledatabase,containingalloriginal materialspreparedaccordingtospecifichistoricalreceiptsand ondocumentedrestorations(thelatterforconsolidantsaddedin moderntimes);
• modest differences in spectral signatures, either related to chemicalsimilarity,ortobroadbandemissioncharacteristicof fluorescenceforsolidsurfaces,stronglyalteredbythe fluores-centsubstrate(e.g.plaster),whichwouldpreventtogetimproved resultsbyhardwareimplementationofhighspectralresolution acquisition.
Duetothementionedlimitations,aquantitativeLIFscanning remoteanalysisseemsnottobefeasiblealsoinperspective. Con-versely,theadvantagesofferedbyfastandsemi-automaticanalysis
capabletorecognizeanomaliesinthedistributionofsurface com-ponents(eitherconsolidants,binders,orevensomepigments)are of interest to theconservator who is aware where to proceed with accurate point analyses utilizing either different spectro-scopic(Raman,LIBS),Spectrometric(SIMS),orimaging(IR,XRF) techniquestohavedefinitematerialidentificationandrespective quantitativeanalysis,thelattersuitableforrenormalizationofLIF scanningintensitydistributiononremotelycollectedimages.
References
[1]D.Anglos,M.Solomidou,I.Zergioti,V.Zaffiropulos,T.G.Papazoglou,C.Fotakis, Laser-inducedfluorescenceinartworkdiagnostics:anapplicationinpigment analysis,Appl.Spectrosc.50(1996)1331–1334.
[2]D.Lognoli,G.Lamenti,L.Pantani,D.Tirelli,P.Tiano,L.Tomaselli,Detectionand characterizationofbiodeteriogensonstoneculturalheritagesbyfluorescence lidar,Appl.Opt.41(2002)1780–1787.
[3]J.Striová,G.Coccolini,S.Micheli,C.Lofrumento,M.Galeotti,A.Cagnini,E. Castellucci,Non-destructiveandnon-invasiveanalysesshedlightonthe real-izationtechniqueofancientpolychromeprints,Spectrochim.ActaA73(2009) 539–545.
[4]G.Cecchi,L.Pantani,V.Raimondi,L.Tomaselli,G.Lamenti,P.Tiano,R.Chiari, Fluorescencelidartechniquefortheremotesensingofstonemonuments,J. Cult.Herit.1(2000)29–36.
[5]D.Comelli,C.D’Andrea,G.Valentini,R.Cubeddu,C.Colombo,L.Toniolo, Fluo-rescencelifetimeimagingandspectroscopyastoolsfornondestructiveanalysis ofworksofart,Appl.Opt.43(2004)2175–2183.
[6]F.Colao,R.Fantoni,L.Fiorani,A.Palucci,Fluorescence-basedportabledevice basedforthespatialscanofsurfaces,inparticularintheareaofCultural Her-itage,ItalianpatentRM2007A000278depositedon21.05.2007.
[7]L.Caneve,F.Colao,L.Fiorani,A.Palucci,Portablelaserradarsystemforremote surfacediagnostics,ItalianpatentRM2010A000606depositedon17.11.2010. [8]I.Borgia,R.Fantoni,C.Flamini,T.DiPalma,A.GiardiniGuidoni,A.Mele,
Lumi-nescencefrompigmentsandresinsforoilpaintingsinducedbylaserexcitation, Appl.Surf.Sc.127(1998)95–100.
[9]F.Colao,R.Fantoni,L.Fiorani,A.Palucci,Applicationofascanninghyperspectral lidarfluorosensortofrescodiagnosticsduringtheCULTURE2000campaignin Bucovina,RevistaMonumentelorIstorice/ReviewofHistoricalMonumentsn. LXXV/1-2(Bucharest,2006)53–61.
[10]F.Colao,L.Caneve,A.Palucci,R.Fantoni,L.Fiorani,Scanninghyperspectrallidar fluorosensorforfrescodiagnosticsinlaboratoryandfieldcampaigns,in:J.Ruiz, R.Radvan,M.Oujja,M.Castillejo,P.Moreno(Eds.),LasersintheConservation ofArtworks(LACONAVII),CRCPress,2008,pp.149–155.
[11]L.Caneve,F.Colao,R.Fantoni,L.Fornarini,Laserinducedfluorescence anal-ysisofacrylicresinsusedinconservationofculturalheritage.Proceedings ofOSAV’2008,The2ndInt.TopicalMeetingonOpticalSensingandArtificial Vision,St.Petersburg,Russia,2008,pp.57–63.
[12]A.C.Rencher,MethodsofMultivariateAnalysis,WileyInterscience,NewYork, 2002.
[13]G.Girouard,A.Bannari,A.ElHartiandA.Desrochers,ValidatedSpectralAngle MapperAlgorithmforGeologicalMapping:ComparativeStudybetween Quick-birdandLandsat-TM,PresentedattheXXthISPRSCongress,Geo-Imagery BridgingContinents,Istanbul,Turkey,12–23July2004.
[14]C.Miliani,M.Ombelli,A.Morresi,A.Romani,Spectroscopicstudyofacrylic resinsinsolidmatrices,Surf.CoatingsTechnol.151–152(2002)276–280. [15]S.Valadas,A.Candeias,J.Mirao,D.Tavares,J.Coroado,R.Simon,A.S.Silva,M.Gil,