Fluorescence LIDAR Technique for Cultural Heritage
Valentina Raimondi, Giovanna Cecchi, David Longnoli, Lorenzo Palombi, Gaia Ballerini
Institute for Applied Physics ‘Nello
The fluorescence lidar technique is a remote sensing technique that enables to analyse a remote object without direct contact with the object itself. This technique is non-destructive and essentially consists in analysing from a distance the light emitted (fluorescence) by the object when the latter is illuminated with a light pulse of a proper colour (i.e. a proper wavelength). The use of fluorescence to point out features that are normally not visible under natural illumination is well known: credit cards are exposed to UV illumination to check their authenticity and paintings are analysed under UV lamps to detect, for example, remnants of transparent varnish layers. Thus, the fluorescence lidar technique enables to apply the fluorescence technique in the outdoor environment by using a pulsed laser to illuminate the target and a telescope to collect the fluorescence light emitted by the target itself.
The fluorescence lidar technique has been applied to the investigation of the cultural heritage only quite recently. Several experiments conducted in the last decade have proved this technique to be a useful tool for the remote diagnostics of monuments, providing helpful information for the assessment of the conservation status of monuments in the outdoor and for the characterization of the materials employed in their construction. Main applications include the detection and characterization of different stones, mortars and other construction materials, of protective treatments, of biodeteriogens and the study of the effects of biocide treatments.
The fluorescence lidar technique does not require sampling procedures and the subsequent laboratory analysis, procedures which can require considerable resources in terms of time, money and specialised personnel, especially in the case of an extensive control of the historical building heritage. Moreover, the acquisition of samples implies a violation of the integrity of the monument and a manipulation of the original material that may lead to artefacts in the results, e.g. stone samples with spores or quiescent forms that become active only once isolated and cultivated in the laboratory. In addition, it allows the control of areas that are difficult to reach without scaffolding since fluorescence measurements are obtained from a remote location.
In the following the fluorescence lidar measurement concept will be briefly introduced together with main applications and some key experiments carried out for the diagnostics of the monuments in the outdoor.
The lidar, acronym of LIght Detection And Ranging, is an instrument that allows the remote analysis of a target by means of a pulsed laser. The lidar can be regarded as the optical counterpart of the radar: the radar operates in the microwave-range, the lidar operates in the visible range of the electromagnetic (e.m.) spectrum and uses a pulsed laser as an excitation source. In general, the lidar allows obtaining information about the geometrical or the chemical-physical characteristics of a target, depending if either the temporal behaviour or the spectral composition (different colours) of the signal is analysed, respectively.
Fluorescence lidar remote sensing is a lidar technique that enables to detect remotely the light emitted (fluorescence, see next section) from the target when the latter is illuminated with a laser of a proper wavelength. A fluorescence lidar (Fig. 3.6.1) is essentially composed of a laser, a telescope, a dispersion system and a detector. The laser beam, which can also be collimated by using a beam expander, is sent to the target and interacts with its internal electronic structure. The fluorescence emitted by the target, which contains information about the chemical-physical characteristics of the target, is collected by the telescope and then fed to the dispersion and detection system, usually featuring high spectral resolution to obtain information, similar to that obtained in the laboratory, directly in the field. The data are finally stored in a PC for the analysis of the signal.
The lidar technique was initially developed for the remote sensing of the atmosphere at the beginning of the 1960s: first papers [1, 2] on this subject were published in 1963, only one year after the invention of the giant pulse laser . In the following years the lidar technique was also applied to the remote sensing of natural waters  leading to the first experiments with fluorescence lidars [5, 6] that were initially used for oil spills detection and characterization and then for the study of dissolved organic matter and phytoplankton. In the 80’s fluorescence lidar remote sensing extended its field of application to the investigation of vegetation fluorescence , with an emphasis on the monitoring of physiological processes and the effect of environmental stresses on plants . At the same time, the development of these applications and further technological advances made it possible to implement more compact and versatile prototypes which could fit the application of the remote sensing technique to the cultural heritage.
was not until the mid 90s that the fluorescence lidar technique was applied
to the diagnostics of the surfaces of historical buildings. First field
experiments [9, 10] for the application of the fluorescence lidar technique
to the monitoring of the cultural heritage were conducted on the Baptistery
and Cathedral of
In the following years further experiments investigated the fluorescence lidar remote sensing potentials for the characterisation of different types of stones [11, 12] and for the detection and characterisation of photoautotrophic biodeteriogens (green algae and cyanobacteria) on stone materials . Other studies were conducted in the laboratory for the characterisation of protective treatments on stone samples  and for the investigation of the effects of biocide treatments .
further step forward was taken in 1997 when the first experiment of lidar
multispectral imaging was performed . The first experiment of
fluorescence imaging concerned the
More recently, a further work  has extended the results obtained previously  with the specific purpose of enhancing the feasibility of identifying different lithotypes on a composite surface and of exploiting the potential of the lidar multispectral imaging technique for an extensive monitoring. The experiment was conducted on the same monuments chosen as targets for the first lidar experiment  held in 1994. During the experiment an extensive fluorescence multispectral imaging of both the monuments was performed, with an emphasis on the monitoring of protective treatments and decorative pigments, features that were not investigated before on monuments with the lidar technique.
Fluorescence is the emission of electro magnetic radiation due to electric dipole transitions between different energy states according to selection rules which reflect the conservation of the angular momentum for the molecule system . Fig. 3.6.2 schematically shows the fluorescence process for a molecule. The molecule is initially in the electronic ground state E0 (at room temperature most molecules are in the lowest vibrational sublevel due to the Boltzmann energy distribution). On the absorption of a photon hnin the molecule is excited to a rotovibrational state E3 of the first excited electronic state. Within this electronic level the molecule decays quickly to the lowest rotovibrational sublevel E3 by means of non-radiative processes. The lowest rotovibrational state has a lifetime that is longer (10-9 s) than those of the upper levels. The transition from the E2 level to the E1 level is associated to the emission of a photon (hnout). Finally, the molecule relaxes in the ground state by non-radiative processes.
Fluorescence in natural stones is due to diadochic and adsorbed cations and to defects in the mineralogical periodicity. The presence of different contributions to fluorescence in the same sample and the effects of sensitisation and quenching, due to other types of cations or different minerals, make the investigation of the fluorescence features fairly difficult. Nevertheless, these features constitute a good source of information about the material under investigation, to be related to both its geochemical and petrographic characteristics. Typical activators in natural samples are divalent or trivalent transition-metal ions and trivalent rare-earth ions . Mineral defect fluorescence usually features a broad emission band in the blue . The most important quenchers in carbonates and oxide minerals are Fe2+, Fe3+, Co2+, and Ni2+, while, for example, Ce3+ has been found to act as a sensitiser, and also as an activator, in calcite [21, 22].
Fluorescence in biodeteriogens is mainly due to pigments and other compounds such as waxes, ferulic acids, phenylpropanoids etc. . Chlorophyll a, that is present in all photoautotrophic organisms, has a typical fluorescence peak around 680 nm. In addition to Chlorophyll a, several biodeteriogens contain other fluorescent pigments, such as chlorophyll b, c and d, and phycobilins. Thus their fluorescent features can be used as a signature to identify the biodeteriogens . For example, green algae and cyanobacteria, which are two taxa of biodeteriogens, show significant differences in their pigment content. The green algae contain Chlorophyll a and small quantities of other pigments, but they do not have phycobiliproteins; cyanobacteria contain Chlorophyll a and other fluorescent pigments including phycobiliproteins, such as phycoerythrin and phycocyanin. The latter ones show typical fluorescence features that can be exploited to differentiate the two taxa on the basis of their Laser Induced Fluorescence (LIF) spectral signatures.
A remarkable enhancement in the readability of the final data can be achieved by using the hyper-spectral fluorescence lidar imaging technique. This essentially consists in scanning the target area with the lidar system and acquiring a complete spectrum for each position so that, at the end of the scansion, a fluorescence hyper-spectral fluorescence image of the investigated area can be retrieved. The measurement concept is synthetically sketched in Fig. 3.6.3: the investigated area on the surface of the monument is ideally divided into m x n 'squares' (Fig. 3.6.3A) and then scanned with the lidar system to obtain a fluorescence image of the target. For each 'square' of the target image, or ‘image pixel’, a full high resolution fluorescence spectrum is acquired (Fig. 3.6.3B). The m x n set of fluorescence spectra is finally analysed to obtain thematic maps that outline specific fluorescence features of the target (Fig. 3.6.3C).
Fluorescence-based thematic maps are particularly attractive for the control of monuments: firstly, they provide a comprehensive assessment on the status of the whole monument and a spatial definition that cannot be obtained by means of mere sampling. Moreover, the opportunity of recording time-dependent, repetitive fluorescence images opens new prospects for reliable control, repeated in time, of the status changes of the monument. Another important aspect of thematic maps is that they make it easier to transfer the knowledge gained with a sophisticated analysis of the fluorescence data, which necessarily requires a specific scientific background, to the conservation specialist and/or to the decision maker.
The hyper-spectral image analysis and then the realisation of thematic maps can be achieved in different ways: one of the simplest ways is to operate a ratio between the areas of two different spectral bands. The area ratio is a useful technique to investigate well known characteristics: for example, the determination of areas affected by biodeteriogens growth can be often achieved by operating a ratio between the spectral range centred around 680 nm, corresponding to the chlorophyll a fluorescence signature, and another suitable spectral range chosen as a background.
A more refined technique for the analysis of hyper-spectral fluorescence data is the Principal Component Analysis (PCA) . Differently from the ratio-based technique, the PCA technique permits a classification without preliminary information about spectral differences of the samples. PCA is a multivariate technique that consists in projecting the spectral information onto a lower dimensional subspace. The latter is built up by a set of orthonormal eigenvectors of the covariance matrix (PCs), chosen so that the first PC describes as much as possible of the covariance among the samples, the second PC as much of the residual covariance and so on. New PCs are added until the residual is considered as just noise. If a sufficient number of PCs are used, each spectrum can be reconstructed by a linear combination of the PCs. The expansion coefficients are called "scores".
Thematic maps can be obtained by plotting the scores of the relevant PC, or a ratio between the scores of two PCs, in a false colour set so as to point out specific features of the samples. It is possible to obtain a thematic map that shows the scores on three different PC using a RGB colour map. The map is composed by three independent monochromatic maps, the red map that shows the scores on first chosen PC, the green one that shows the scores on second chosen PC, the blue one that shows the scores on third chosen PC. The three maps are then summed up in RGB false colour image.
3.6.4 shows a photo of miscellaneous samples as they were arranged during a
joint experiment aimed at the investigation of the fluorescence features of
natural stones and other building materials . The samples were arranged
on a special scaffold that was previously covered with non fluorescent
material to simulate an operative scenario. The samples included several
types of marbles, among which
Thematic maps obtained either operating ratios between proper spectral bands or applying the PCA technique are presented in Fig. 3.6.4. The thematic map of Fig. 3.6.4 was obtained by applying the PCA technique in the spectral range 355-387 nm to the spectra collected with the 300 nm excitation wavelength. The spectra were normalized to the maximum value before the PCA. The colour map was built by superimposing three independent channels Red, Green and Blue. The Red channel was associated to the PC1 score, the Green channel to the PC2 score, the Blue channel to the PC3 score. The map permits to distinguish the marbles (MS1 and MS2 in green colours and MCA in blue colours) and the sandstones (red colours). Lime mortar as well is in red colour, but it can be differentiated from sandstones with a PCA applied to different spectral ranges. Similar thematic maps can be obtained by applying the PCA in different spectral ranges allowing the differentiation of pozzolanic mortar from the other materials, as well as for the differentiation among similar limestone lithotypes (calcirudites and packstones).
The use of conservative treatments on stony materials has old origins: however, often there is no documented trace of the type of treatment applied in the past times on certain monuments. Sometimes even modern treatments can be craft-made. Consequently, a method able to detect them on the monument’s surface, and possibly to identify them, can be extremely helpful.
Fig. 3.6.5 shows the comparison between the fluorescence spectra of a sample of dolomitic marble and of samples of the same material with different protective treatments as shown in the figure’s legend. Although the spectral shapes are quite similar, a more refined analysis with the PCA technique allowed to differentiate the samples treated with different treatments and the control sample. Fig. 3.6.6 shows the results obtained processing the LIF spectra with the PCA: a first analysis to the full data set enables to distinguish the samples with the protective treatments from the control one. Fig. 3.6.6a shows the PC4 score plotted against the PC1 one to differentiate the Fomblin (FOM) and the Rinforzante OH (ROH) protective treatments. Fig. 3.6.6b shows the PC2 score plotted against the PC1 one to distinguish also the other two protective treatments, Akeogard (AKCO) and Paraloid B72 (PB72), from the control sample.
The characterization of photoautotrophic biodeteriogens with the fluorescence lidar technique is based on the analysis of the fluorescence signatures of the chlorophyll and the other fluorescent pigments. Fig. 3.6.7 shows as an example the fluorescence spectra of several species belonging to different classes. These spectra were obtained in the laboratory with excitation at 514 nm during a set of measurements aimed at the identification of phytoplankton in the sea, but the method can be equally applied to the identification of biodeteriogens on monuments.
Specific LIF measurements were also conducted on some species of cyanobacteria and green algae isolated directly on monuments . The species Pleurococcus sp. and Chroococcus sp. belonging to the green algae and cyanobacteria, respectively, were cultivated in the laboratory and then inoculated at different concentrations on marble (dolomitic marble) and sandstone (Pietra di Lecce) samples.
Fig. 3.6.8 shows the fluorescence spectra obtained on marble samples inoculated with different biodeteriogens: Fig. 3.6.8a shows the fluorescence spectrum of the sample inoculated with green algae (Pleurococcus sp.) and Fig. 3.6.8b the fluorescence spectrum of that inoculated with cyanobacteria (Chroococcus sp.). The concentration was 2.000 cells/cm2. Excitation wavelength was 488 nm. The comparison between the two spectra points out the remarkable differences in the spectral shape: the green algae show the typical Chlorophyll a fluorescence peak at 680 nm, whereas the cyanobacteria, in addition to the chlorophyll fluorescence peak, have additional peaks at 660 nm and 570 nm which are typical of phycocyanin and phycoerithrin, respectively. The minimum concentrations detected in the laboratory were 200 cells/cm2, while lidar measurements in the outdoor from a distance of about 20 m detected only higher concentrations (20,000 cells/cm2) that were anyway not detectable by the naked eye.
September 2000 a joint experiment was performed on the Cathedral and the
Cathedral and the Baptistery of
During the experiment, 12 areas of the two monuments were monitored: 8 areas on the Cathedral and 4 areas on the Baptistery. The locations of the monitored areas are shown in Fig. 3.6.9a and in Fig. 3.6.9b. The investigated areas were scanned with the LTH lidar system to acquire hyper-spectral fluorescence images of the target. In addition, several point fluorescence measurements were acquired on spots selected for their special features with both the lidar systems.
The experiment allowed pointing out the presence of an unknown protective treatment on various areas of the Cathedral façade, especially around the Protiro, that were not visible by the naked eye. In addition, several thematic maps referring to areas of both the Cathedral and the Baptistery showed the presence of biological growth. Several pieces of information were instead retrieved concerning the lithotypes and other materials employed for the decoration of the monuments . Here only a selection of the main results are presented.
Fig. 3.6.10 shows the fluorescence measurements taken in the area around the Protiro with the FLIDAR system featuring an excimer laser at 308 nm as an excitation source. Fig. 3.6.10a displays the spots where the point fluorescence measurements were taken. The results obtained exciting at 308 nm are shown in Fig. 3.6.10b: the fluorescence signature of residual protective treatment is present in almost all the spectra (spot 3 and 4 mainly show the fluorescence due to the stone). The 308-nm excitation in the areas affected by the protective treatment yields a peak at about 380 nm. The presence of residual protective treatment is more marked in the area around the Protiro of the Cathedral, especially in those areas that are less subject to weathering.
A picture of the balustrade (area L in Fig. 3.6.9b) on the roof of the Baptistery is shown in Fig. 3.6.11a. Fig. 3.6.11b shows a thematic map obtained with the 355 nm excitation: here the ratio between the PC2 and PC1 scores is plotted. The yellow-red pixels point out the areas characterised by the presence of biodeteriogens. The shape of the PC2, actually, is associated to the chlorophyll fluorescence spectral shape while the shape of PC1 is clearly related to the fluorescence of the stone substrate (not shown here). In this image the laser beam is not normal to the scanned surface due to the large pointing angle. This is a clear example of an area difficult to be reach without scaffolding for a monitoring of biological growth on a regular basis.
The fluorescence lidar technique has been applied successfully to the diagnostics of monuments in the outdoor. The technique allows to investigate several aspects concerning the conservation status of the monument and to obtain information on the materials employed for its construction that can be used for a correct arrangement of restoration intervention or for the reconstruction of its history. In particular the technique was demonstrated helpful for the detection and characterization of biodeteriogens and protective treatments. Further experiments pointed out the feasibility of a fluorescence-based characterization of different lithotypes and other building materials such as mortars, cements. This information can be exploited to identify areas of previous (unknown) restorations/interventions or to rebuild the construction history of the monument.
Several experiments on monuments have also demonstrated the practicability of extensive monitoring of historical buildings with thematic maps by using the hyper-spectral fluorescence lidar imaging technique. In particular, thematic maps can be used to outline the presence of the areas subject to protective treatment and to biological growth, to identify areas constituted by different lithotypes. Thematic maps can be seen as a helpful support for decision-makers in view of a routinely planned monitoring of our stone cultural heritage.
The authors wish to remember the
memory of Prof. Luca Pantani, prematurely died in 2001, who first believed in
this application of the fluorescence lidar technique to the investigation of
the cultural heritage and undertook this activity with great enthusiasm. The
authors are very grateful to Prof. Sune Svanberg and his group of the Lund
Institute of Technology,
 G. Fiocco, L.D. Smullin: Detection of Scattering Layers in the Upper Atmosphere (60-140 km) by Optical Radar. Nature 199 (1963) 1275-1276.
 M.G.H. Ligda: Proc. Conf. Laser Tecnol., 1st, San Diego, USA (1963) 63-72.
 F.J. Mc.Clung, R.W. Hellwarth: Giant Optical Pulsation from Ruby. J.Appl. Phys. 33 (1962) 828-829.
 G.D. Hickman, J.E. Hogg: Application of an Airborne Pulsed Laser for Near Shore Bathymetric Measurements. Rem. Sens. Environ. 1 (1969) 47-58.
 R.M. Measures, M.
Bristow: The Development of a Laser Fluorosensor for Remote Environmental
 J.F. Fantasia, T.M
Hard., H.C. Ingrao: An Investigation of Oil Fluorescence as a Technique for
Remote Sensing of Oil Spills. Report No.DOT-TSC-USCG-71-7,
 F.E. Hoge, R.N. Swift, J.K. Yungel: Feasibility of Airborne Detection of Laser-Induced Fluorescence Emission from Green Terrestrial Plants. Appl.Opt. 22 (1983) 2991-3000.
 F.E. Hoge: Oceanic
and terrestrial lidar measurements. In: Laser remote chemical analysis
(Measures R.M Ed.), John
 V. Raimondi, L.
Masotti, G. Cecchi, L. Pantani: Remote sensing of cultural heritage: a new
field for lidar fluorosensors. Proc. of 1st International Congress on:
“Science and Technology for the Safeguard of Cultural Heritage in the
 V. Raimondi, G. Cecchi, L. Pantani, R. Chiari: Fluorescence lidar monitoring of historic buildings. Appl. Opt. 37 (1998) 1089-1098.
 G. Cecchi, L. Pantani, V. Raimondi, D. Tirelli, R. Chiari: The fluorescence lidar technique for the remote sensing of stony materials in ancient buildings. SPIE 2960 (1996) 163-171.
 G. Cecchi, L. Pantani, V. Raimondi, L. Tomaselli, G. Lamenti, P. Tiano, R. Chiari: Fluorescence lidar technique for the remote sensing of stone monuments. J. Cult. Heritage 1 (2000) 29-36.
 G. Cecchi, L. Pantani, V. Raimondi, D. Tirelli, L. Tomaselli, G. Lamenti, M. Bosco, P: Tiano: Fluorescence lidar technique for the monitoring of biodeteriogens on the cultural heritage. In: Remote Sensing for Geography, Geology, Land Planning, and Cultural Heritage (D. Arroyo-Bishop et al. Eds) SPIE 2960 (1996) 137-147.
 G. Ballerini, S. Bracci, L.
Pantani, P. Tiano: Lidar remote sensing of stone cultural heritage: detection
of protective treatments. Opt.
 D. Lognoli, G. Lamenti, L.Pantani, D. Tirelli, P. Tiano, L.Tomaselli: Detection and characterisation of biodeteriogens on stone cultural heritage by fluorescence lidar. Appl. Opt. 41 (2002) 1780-1787.
 P. Weibring, T. Johansson, H. Edner, S. Svanberg, B. Sundnér, V. Raimondi, G. Cecchi, L. Pantani: Fluorescence lidar imaging of historical monuments. Appl. Opt.-LP 40 (2001) 6111-6120.
 D. Lognoli, G. Cecchi, I. Mochi, L.
Pantani, V. Raimondi, R. Chiari, Th. Johansson,
P. Weibring, H. Edner, S. Svanberg: Fluorescence Lidar Imaging of the
 S. Svanberg: Atomic and Molecular
Spectroscopy. Springer Verlag,
 H.G. Machel, R.A. Mason, A.N. Mariano, A. Mucci: Causes and emissions of luminescence in Calcite and Dolomite. In: Luminescence microscopy and spectroscopy: qualitative and quantitative applications (Barker C.E. and Kopp O.C. edts) SEPM Short Course 25 (1991) 9-25.
 H. Gies: Activation Possibilities and Geochemical Correlations of Photoluminescing carbonates, Particularly Calcites. Mineral. Deposita (Berl.) 10 (1975) 216-227.
 G.M. Aguilar, M.I. Osendi: Fluorescence of Mn2+ in CaCO3. Journal of Luminescence 27 (1982) 365-375.
 G. Blasse, G.M. Aguilar: Luminescence of natural calcite (CaCO3). Journal of Luminescence 29 (1984) 239-241.
 K.V. Mardia, J.T. Kent, J.M. Bibby:
Multivariate Analisys. Accademic Press,
 Z. G. Cerovic, G. Samson, F. Morales,
 C.S. Yentsch, D.W. Menzel: A method
for the determination of phytoplankton chlorophyll and phaeophytin by fluorescence.
 L. Pantani, G. Cecchi, D. Lognoli,
I. Mochi, V. Raimondi, D. Tirelli, M. Trambusti, G. Valmori, P.Weibring, H.
Edner, T. Johansson, S. Svanberg: Lithotypes characterization with a
fluorescence lidar imaging system using a multi-wavelength excitation source.
In: Remote Sensing for Environmental Monitoring, GIS Applications, and
Geology II, (M. Ehlers Ed.) SPIE Proc. Vol. 4886, SPIE,
 M. Bazzani, G. Cecchi: Algae and Mucillagine Monitoring by Fluorescence Lidar Experiments in Field. EARSeL Advances in Remote Sensing 3 (1995) 90-101.
 B. Breschi, G. Cecchi, L. Pantani, D. Tirelli, G. Valmori: FLIDAR a multipurpose fluorosensor –spectrometer. EARSeL Advances in Remote Sensing 1 (1992) 72-78.
 G. Cecchi, P. Mazzinghi, L.
Pantani, R. Valentini, P. De Angelis: Remote sensing of chlorophyll a
fluorescence of vegetation canopies:
P. Weibring, H. Edner, S. Svanberg: Versatile
Applied Physics Institute ‘Nello Carrara’
Italian National Research Council (CNR-IFAC)
Via Panciatichi 64
Department of Physics
Lund Institute of Technology
Institute for Applied Physics ‘Nello Carrara’
Via Panciatichi 64
Institute for Applied Physics ‘Nello Carrara’
Via Panciatichi 64