Revealing Underdrawings in Wall Paintings of Complex Stratigraphy with a Novel Reflectance Photoacoustic Imaging Prototype

Abstract

Revealing precious hidden features by a completely non-invasive approach is one of the crucial issues in the Heritage Science field. In this regard, concealed fresco paintings still represent an analytical challenge. This paper addresses the specific issue in wall painting diagnostics by the photoacoustic (PA) imaging technique, already proven to be efficient in revealing underdrawings and internal stratigraphy in movable paintings on paper and canvas. A newly set-up reflection PA prototype was applied here for the first time to probe the charcoal, graphite and sinopia hidden sketch drawings in concealed (gypsum, limewash, overpainted) wall paintings. The results presented here push forward the frontiers of the PA imaging technique and point to its potential effectiveness of revealing hidden underdrawings in historical wall paintings with complex stratigraphy.

Full text

Journal of
Imaging
Article
Revealing Underdrawings in Wall Paintings of Complex
Stratigraphy with a Novel Reflectance Photoacoustic
Imaging Prototype
Antonina Chaban 1, *, George J. Tserevelakis 2, Evgenia Klironomou 2, Raffaella Fontana 1, Giannis Zacharakis 2
and Jana Striova 1


Citation: Chaban, A.; Tserevelakis,
G.J.; Klironomou, E.; Fontana, R.;
Zacharakis, G.; Striova, J. Revealing
Underdrawings in Wall Paintings of
Complex Stratigraphy with a Novel
Reflectance Photoacoustic Imaging
Prototype. J. Imaging 2021,7, 250.
https://doi.org/10.3390/
jimaging7120250
Academic Editors: Filippo Stanco
and Guillaume Caron
Received: 29 September 2021
Accepted: 18 November 2021
Published: 24 November 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
1National Institute of Optics INO-CNR, 50125 Florence, Italy; raffaella.fontana@ino.cnr.it (R.F.);
jana.striova@cnr.it (J.S.)
2Institute of Electronic Structure and Laser, Foundation for Research and Technology Hellas,
70013 Heraklion, Crete, Greece; tserevel@iesl.forth.gr (G.J.T.); eklironomou@physics.uoc.gr (E.K.);
zahari@iesl.forth.gr (G.Z.)
*Correspondence: antonina.chaban@ino.cnr.it
Abstract:
Revealing precious hidden features by a completely non-invasive approach is one of the
crucial issues in the Heritage Science field. In this regard, concealed fresco paintings still represent
an analytical challenge. This paper addresses the specific issue in wall painting diagnostics by the
photoacoustic (PA) imaging technique, already proven to be efficient in revealing underdrawings
and internal stratigraphy in movable paintings on paper and canvas. A newly set-up reflection PA
prototype was applied here for the first time to probe the charcoal, graphite and sinopia hidden
sketch drawings in concealed (gypsum, limewash, overpainted) wall paintings. The results presented
here push forward the frontiers of the PA imaging technique and point to its potential effectiveness
of revealing hidden underdrawings in historical wall paintings with complex stratigraphy.
Keywords:
photoacoustic; infrared reflectance; non-invasive; underdrawings; wall paintings;
complex stratigraphy
1. Introduction
Numerous historical wall paintings remain in part or completely covered by sub-
sequent layers, including retouchings, overpaints, mortar coats, etc. A rediscovery of
the hidden features of wall paintings is a frequent and challenging case in conservation.
A reliable
non-invasive diagnostic procedure, as the first step, is crucial to a painting’s
correct characterization and eventual safe uncovering. There is a need for a robust method-
ology that is capable of addressing all the relevant art historical, material and conservation
queries prior to any intervention on such multilayered wall paintings.
Valuable hidden features may include original underdrawing or even an underpaint-
ing, realized in fresco, mezzo fresco or secco techniques. The buon fresco (or true fresco) is
realized by applying the sketch and paint (pigment mixed with water without a binding
agent) on lime mortar while it is still wet. A secco method foresees the pigment application
on a dry surface and therefore requires a binding medium (e.g., egg tempera, oil, glue,
acrylic binder) to attach the pigment to the wall. A mezzo fresco indicates a wall painting
technique where the pigment is applied with water onto the humid (almost dry) mortar,
so that it only slightly penetrates into the wall. Detailed information about wall painting
execution techniques can be found elsewhere [
1
–
5
]. Further interventions could see the
use of limewash coat to conceal the original sketch or the changes of the final appearance
of a painting by retouching (further paint layer applied a secco). Heritage experts face
even such cases where original paintings were covered by subsequent mortar and paint
layers due to changes in taste, religious belief, ownership or alterations of the building.
J. Imaging 2021,7, 250. https://doi.org/10.3390/jimaging7120250 https://www.mdpi.com/journal/jimaging

J. Imaging 2021,7, 250 2 of 12
Indeed, many examples of precious paintings or sketch drawings covered during history
are reported [6–11].
As a matter of fact, many historical paintings are characterized by often unknown
internal stratigraphy. Near-infrared (NIR) reflectance imaging has become the most consol-
idated method in the detection of underdrawings underneath the depicted surface [
12
–
19
].
This method exploits an increased penetrability of near-infrared radiation through a picto-
rial layer (most of artistic pigments have low scattering/absorption in the near-infrared
range) and strong optical absorption of sketch materials (charcoal, iron-gall ink etc.), in
the condition of a strong diffuse near-infrared reflectance from the ground. Many studies
can be found in the literature on NIR imaging used to reveal underdrawings in fresco
and secco wall paintings underneath retouched areas, while little can be found on deeper
concealed features (covered by limewash, concealing mortar layers etc.). A dedicated study
has already proven the efficiency of NIR imaging to reveal iron-gall ink and carbon-based
sketches on mock-ups with a limewash layer up to 200
µ
m thick, by selecting the most
appropriate wavelength [20].
The literature reports studies on infrared thermography that revealed hidden wall
decoration layers, including wall frescoes [
10
]. Further rapidly developing techniques
applied to revealing hidden features in multilayered paint stratigraphy include THz
Imaging [21–24]
, Optical Coherence Tomography [
25
–
27
] and the newly introduced epi-
illumination Photoacoustic (PA) Imaging [
28
]. In terms of spatial resolution and image
contrast, the latter showed a particularly strong potential for the matter of this study. In
previous years, PA imaging was applied to cultural heritage objects only in transmission
mode [
28
–
33
]. Such configuration cannot be used on wall paintings, since the pictorial
layer on its backside is bonded to the structural support (wall, vault etc.). The novel
epi-illumination PA imaging set up proved its efficiency in detecting underdrawings in sim-
ulated multilayered artworks [
28
], pointing to its applicability in a wide range of Cultural
Heritage objects of arbitrary forms and shapes. Here, we tested it on specifically designed
wall painting mock-ups, prepared following the historical methodologies. In particular,
the scope of this work was to evaluate the effectiveness of the new epi-illumination PA
imaging set-up to detect the sketch drawings in simulated wall paintings. These features
are hidden under a secco paint layer or under both the fresco paint and the concealing
limewash or gypsum coats of two different thicknesses.
We compared and complemented here the PA imaging results with the information ob-
tained by research-grade near-infrared reflectance (NIR) imaging, operated in
380–2500 nm
.
The information on the layer thickness was achieved by laser scanning profilometry.
2. Experimental
2.1. Methods
2.1.1. Photoacoustic Imaging Set-Up
The reflection-mode PA imaging apparatus employs a Q-switched Nd:YAG laser emit-
ting infrared radiation at 1064 nm for the excitation of a PA signal. The object is irradiated
from its front side to generate a laser-induced ultrasound from the underlying hidden
sketch regions. The PA waves are transmitted through the overlying layers and water
prior to their detection in reflectance configuration by a broadband spherically focused
piezoelectric transducer. The integration of immersion ultrasonic detection (e.g., here in
a water medium) minimizes acoustic reflection losses at poorly matched interfaces such
as between paint layers and air [
34
,
35
]. The signals are subsequently enhanced by two
low-noise radio frequencies (RF) amplifiers prior digitization and recording of PA wave-
forms by an oscilloscope. The image is formed through raster scanning along the analyzed
surface using a set of high-precision XY motorized stages. The recorded waveforms were
averaged two times for signal-to-noise ratio (SNR) improvement, transferred to a computer,
and band-passed between 100 kHz and 30 MHz for high-frequency noise elimination
before the estimation of the peak-to-peak PA amplitude value, providing the contrast of the
resulting 8-bit images. The pulse energy at the object’s plane surface was kept below
2.3 mJ

J. Imaging 2021,7, 250 3 of 12
(corresponding surface fluence for spot size 1 mm
2
is less than 0.3 J/cm
2
), following the
results of a preliminary investigation of the optimum irradiation parameters (pulse energy,
number of averaging waveforms), required to provide sufficient SNR levels without the
presence of any apparent photodamage effects. Each mock-up was placed at the bottom of
a 3D-printed sample holder filled with distilled water, serving as an immersion medium
for the efficient propagation and subsequent detection of PA signals. The scanned regions
had dimensions ranging between 1.5
×
1.5 to 2.5
×
2.5 cm
2
and were sampled using
a pixel
size of 300
×
300
µ
m
2
. The total time required for the recording of a PA image ranged
from 40 min to 2 h. Control and synchronization of the PA imaging system was carried
out by means of a custom-developed software, while image-processing operations were
performed in ImageJ and MATLAB programming environment. The pixel size of all PA
images presented in this study is 300
×
300
µ
m
2
. Further technical details on the PA
imaging set-up can be found in [28].
2.1.2. Visible-Near-Infared (VIS-NIR) Reflectance Imaging
The multispectral VIS-NIR scanner developed at the National Institute of Optics
acquires the backscattered radiation from the measured surface. It scans with a spatial
sampling of 4 point/mm (250
µ
m), generating 32 monochromatic images in the visible
(from 380 to 780 nm) and in the infrared region (790 to 2500 nm) with a spectral sampling
step of 20–30 and 50–100 nm, respectively [
36
,
37
]. The VIS-NIR multispectral scanner
illuminates the surface and collects the backscattered radiation in a point-by-point modality.
The optical head bears the lighting system and the collecting. The instrument allows to
scan continuously an area of up to 1 m
2
within 3 h, exploiting two orthogonally mounted
high-precision linear-translation stages, equipped with optical encoders. The acquisition
time for a single point is about 1 msec. The constant motion of the optical head prevents
the surface of the painting from being heated significantly. It works at a distance of
12 cm
from the object’s surface in a 45
◦
/0
◦
(illumination/detection) configuration according
to CIE standards. The system is operated through a custom-developed software, with
a simultaneous
control of movement of scanner head, autofocus and images acquisition.
The pixel size of all NIR images presented in this study is 250 ×250 µm2.
Processing in ImageJ was applied first to enhance the PA result. Subsequently,
the same image processing function was applied to NIR images for comparison of the
experimental results.
2.1.3. Conoscopic Microprofilometry
A conoscopic microprofilometer acquires a collection of vertices over a regular grid,
creating a 3D model of the measured surface. The working distance from the surface is
4 cm
, with a dynamic of 8 mm (depth of field) whereas the acquisition frequency is about
400 points/s. The maximum scanning area is 30
×
30 cm with 1
µ
m axial and
20 µm
lateral
resolution [
38
,
39
]. It is operated through a custom-developed software. Laser scanning
conoscopic microprofilometry provided a topographic map of the mock-up’s surface,
supplying height (z) values, which were used here for thickness and surface topography
measurements. The time required for the recording of an area 12
×
12 cm was of 1 h
45 min
.
The thickness of hiding layers was calculated through subtraction of 3D topography maps
acquired on the mock-ups before and after application of painting layers and coats.
2.2. Experimental Mock-Ups
The effectiveness of the PA imaging technique was evaluated on an ad-hoc prepared
mock-ups simulating a real wall painting. In particular, the analyzed hidden features and
top layers are described in Table 1. The mock-ups stratigraphy is illustrated in Figure 1.

J. Imaging 2021,7, 250 4 of 12
Table 1. Analyzed mock-ups.
Hidden Features Hiding Layer Sample Code
Material Measured Thickness
Charcoal and graphite tempera EB 1paint 40–70 µm 1
Charcoal and sinopia 1 layer gypsum + glue 60–80 µm2
2 layers gypsum + glue 130–190 µm
Charcoal and sinopia +fresco paint EB 11 layer gypsum + glue 60–80 µm 3
Charcoal and sinopia 1 layer limewash + milk 60–80 µm
4
2 layers of limewash + milk 90–140 µm
Charcoal + fresco paint RS 21 layer limewash + milk 60–80 µm 5
1EB: Egyptian blue pigment; 2RS: raw sienna pigment.
J. Imaging 2021, 7, x FOR PEER REVIEW 5 of 12
Figure 1. Stratigraphy scheme of the wall painting mock-ups: (a) Sample 1, containing underdraw-
ings hidden under Egyptian blue secco paint layer; (b) Sample 2, which contains underdrawings,
covered by 1 and 2 gypsum coats, and Sample 3, which contains underdrawings, covered by fresco
and by 1 gypsum coat. (c) Sample 4, which contains underdrawings, covered by 1 and 2 limewash
coats, and Sample 5, which contains underdrawings, covered by fresco and 1 limewash coat.
Figure 1.
Stratigraphy scheme of the wall painting mock-ups: (
a
) Sample 1, containing underdraw-
ings hidden under Egyptian blue secco paint layer; (
b
) Sample 2, which contains underdrawings,
covered by 1 and 2 gypsum coats, and Sample 3, which contains underdrawings, covered by fresco
and by 1 gypsum coat. (
c
) Sample 4, which contains underdrawings, covered by 1 and 2 limewash
coats, and Sample 5, which contains underdrawings, covered by fresco and 1 limewash coat.

J. Imaging 2021,7, 250 5 of 12
The mock-ups were prepared according to the traditional recipe of true fresco
technique [1,2]
, within the premises of Accademia dell’Affresco in Padua (Italy). After
1 year
of natural carbonatation, all the hiding layers (secco paint, gypsum and lime-
wash) were applied onto the mock-ups’ surfaces at the Opificio delle Pietre Dure in
Florence (Italy).
Fresco mock-ups: a thick layer (1 cm) of medium coarse mortar called arriccio (a mix
of slaked lime and medium grain sand (1:2) with water) was applied on a lightweight
wood-fiber support. Subsequently, a thin layer (2–3 mm) of fine coarse mortar called
intonachino (a mix of slaked lime and fine grain sand (1:2) with water) was applied upon it.
Outline/sketch drawings: In order to simulate the characteristic fresco underdrawings,
a black
charcoal pigment nero carbone and sinopia (both Dolci, Verona) were applied on the
lime mortar while it was still wet. The sketch was transferred to the mortar surface using the
traditional for fresco paintings spolvero (pouncing) technique. Additional graphite sketches
(Koh-I-Noor Hardtmuth pencil, hardness level 2B) were realized on the dry
mortar surface
.
Pigments: Two characteristic for fresco (stable in alkaline environment) pigments—
Egyptian blue (EB) and raw sienna (RS) [
40
]—were applied on a fresh mortar, shortly
after sketch.
Hiding coats: EB was applied a secco using egg yolk tempera binder (egg yolk and
distilled water, 1:1) on the dry mortar surface to simulate simple retouching or overpaint
(Sample 1). Gypsum (calcium sulphate dihydrate CaSO
4
. 2H
2
O, bound with rabbit glue
dissolved in water in proportion 1 g:12 mL) and limewash (lime, water with a small addition
of milk) coats were applied on dry substrate to conceal the underdrawings, respectively, in
Samples 2 and 3 and in Samples 4 and 5. Gypsum and limewash coats were applied by
brush in one and two layers to obtain two different thicknesses as specified in Table 1.
3. Results
3.1. Imaging Charcoal and Graphite Outlines under Tempera Paint: Sample 1
A graphite drawing representing a city skyline as well as the charcoal line (Figure 2A)
were hidden under a 70-micron thick Egyptian blue paint layer, applied with egg tempera
binder (Figure 2B). The raw PA and NIR images were processed in equal way by the means
of ImageJ, applying 1.0% contrast stretching. The PA image (Figure 2C) of the painted
sample, obtained in reflection mode, clearly reveals the sketch drawing made by graphite
pencil directly on the dried mortar. Both graphite and charcoal traits are revealed in the
NIR reflectogram (Figure 2D). However, graphite has a significantly higher absorption than
charcoal at 1064 nm [
29
,
41
,
42
]. Since the image contrast in PA image is relative per scanned
area, the charcoal line presents in this specific case a too low contrast to be visible in the
PA image. For revealing low contrast details in a PA image, application of post-processing
algorithms or additional area scans might be needed. These tests lie beyond the aims of the
first experimentation and are expected to be addressed by future studies.
J. Imaging 2021, 7, x FOR PEER REVIEW 6 of 12
3. Results
3.1. Imaging Charcoal and Graphite Outlines under Tempera Paint: Sample 1
A graphite drawing representing a city skyline as well as the charcoal line (Figure
2A) were hidden under a 70-micron thick Egyptian blue paint layer, applied with egg
tempera binder (Figure 2B). The raw PA and NIR images were processed in equal way by
the means of ImageJ, applying 1.0% contrast stretching. The PA image (Figure 2C) of the
painted sample, obtained in reflection mode, clearly reveals the sketch drawing made by
graphite pencil directly on the dried mortar. Both graphite and charcoal traits are revealed
in the NIR reflectogram (Figure 2D). However, graphite has a significantly higher absorp-
tion than charcoal at 1064 nm [29,41,42]. Since the image contrast in PA image is relative
per scanned area, the charcoal line presents in this specific case a too low contrast to be
visible in the PA image. For revealing low contrast details in a PA image, application of
post-processing algorithms or additional area scans might be needed. These tests lie be-
yond the aims of the first experimentation and are expected to be addressed by future
studies.
Figure 2. The charcoal (line in upper left corner) and graphite (city skyline) traits on the mortar: visible images (A) prior
to paint application and (B) after Egyptian blue tempera application (secco overpaint); (C) a PA image and (D) NIR reflec-
tance image of (B). The scale bar in (A) applies to all the images.
3.2. Imaging Charcoal and Sinopia Drawings under Gypsum Coat: Samples 2 and 3
Visible images of Sample 2 are shown in Figure 3 along with the relevant representa-
tive results. In top raw, sinopia and charcoal traits on substrate (Figure 3A) are hidden by
a 80-micron thick gypsum layer (Figure 3B); in bottom raw, mixture of sinopia and char-
coal is covered by 190-micron thick gypsum layers (Figure 3B). Both PA and NIR images
were processed in an equal way by the means of ImageJ. The function of 1.0% contrast
stretching and 0.6 gamma correction were applied to the PA and NIR images in the top
raw and 1.0% contrast stretching with a min value of 10.826 to the PA and NIR images in
the bottom raw. The min value corresponds to the lower limit of the display range for the
processed 16-bit images (brightness range: 0–65.536), so that all pixels with a brightness
equal to or below 10.826 were set to zero (total black). The representative images of the
PA and NIR investigations are displayed respectively in Figure 3C and Figure 3D,E. For
the NIR investigation, first, the reflectogram centered at 1050 nm (1000–1100 nm) is shown
in Figure 3D due to the overlap with the excitation wavelength exploited by PA. Then, the
last column shows the highest contrast NIR reflectogram, at 950 nm and at 1292 nm, for
the two areas, respectively. As for the results shown in the top raw, the charcoal line is
clearly distinguished by both PA and NIR methods, while sinopia only by NIR (Figure
3C–E).
The PA technique demonstrates a good capability in revealing charcoal traits hidden
under the gypsum coats (Figure 3C). In the top raw (Figure 3), a 80-micron thick gypsum
layer covers both charcoal and sinopia lines. A charcoal line hidden under one and two
Figure 2.
The charcoal (line in upper left corner) and graphite (city skyline) traits on the mortar: visible images (
A
) prior to
paint application and (
B
) after Egyptian blue tempera application (secco overpaint); (
C
) a PA image and (
D
) NIR reflectance
image of (B). The scale bar in (A) applies to all the images.

J. Imaging 2021,7, 250 6 of 12
3.2. Imaging Charcoal and Sinopia Drawings under Gypsum Coat: Samples 2 and 3
Visible images of Sample 2 are shown in Figure 3along with the relevant representative
results. In top raw, sinopia and charcoal traits on substrate (Figure 3A) are hidden by
a 80-micron
thick gypsum layer (Figure 3B); in bottom raw, mixture of sinopia and charcoal
is covered by 190-micron thick gypsum layers (Figure 3B). Both PA and NIR images were
processed in an equal way by the means of ImageJ. The function of 1.0% contrast stretching
and 0.6 gamma correction were applied to the PA and NIR images in the top raw and 1.0%
contrast stretching with a min value of 10.826 to the PA and NIR images in the bottom
raw. The min value corresponds to the lower limit of the display range for the processed
16-bit images (brightness range: 0–65.536), so that all pixels with a brightness equal to
or below 10.826 were set to zero (total black). The representative images of the PA and
NIR investigations are displayed respectively in Figure 3C and Figure 3D,E. For the NIR
investigation, first, the reflectogram centered at 1050 nm (1000–1100 nm) is shown in
Figure 3D due to the overlap with the excitation wavelength exploited by PA. Then, the
last column shows the highest contrast NIR reflectogram, at 950 nm and at 1292 nm, for the
two areas, respectively. As for the results shown in the top raw, the charcoal line is clearly
distinguished by both PA and NIR methods, while sinopia only by NIR (Figure 3C–E).
J. Imaging 2021, 7, x FOR PEER REVIEW 7 of 12
layers of gypsum cover (respectively 80 and 190 μm thick) is detectable by the PA tech-
nique, as shown in Figure 3C. This method presents a good capability in revealing char-
coal under the gypsum coats and even allows clear detectability of charcoal in lower con-
centration (mixed with sinopia) under two layers of gypsum (thickness 190 μm). At the
same time, sinopia is not detectable by photoacoustic imaging at the exploited wavelength
(1064 nm) but is detectable by near-infrared reflectance imaging. The combined PA imag-
ing and NIR reflectance imaging approach is shown to be helpful here in the detection of
both underdrawing materials (charcoal and sinopia) and making a hypothesis about its
type.
Figure 3. Revealing hidden sketch drawings under gypsum coat in Sample 2: (A) visible image prior to gypsum applica-
tion; (B) visible image after gypsum coat application; (C) PA imaging result; (D,E) NIR reflectance images. The scale bar
in (A) is valid for all the images of the row.
Visible images of Sample 3 (Figure 4A–C) show yet another simulated scenario in
which the underdrawing traits, made with sinopia and with a charcoal/sinopia mixture,
are covered first with an Egyptian blue fresco paint (Figure 4B) and then with 80-micron
thick gypsum layer (Figure 4C). Both PA and NIR images were processed in equal way
by means of ImageJ, applying 1.0 contrast stretching with min value: 10.826. The PA image
of the stratified sample, Figure 4D, reveals the mixed charcoal/sinopia trait whereas the
sinopia lines are visible in NIR reflectograms (Figure 4E,F).
Figure 4. Revealing hidden sketch drawings under fresco paint and gypsum coat in Sample 3: visible images of (A) initial
drawing on the substrate; (B) after fresco paint application; (C) after gypsum coat application; (D) PA image of (C); (E,F)
NIR images of (C). The scale bar in (A) is valid for all the images.
Figure 3.
Revealing hidden sketch drawings under gypsum coat in Sample 2: (
A
) visible image prior to gypsum application;
(
B
) visible image after gypsum coat application; (
C
) PA imaging result; (
D
,
E
) NIR reflectance images. The scale bar in (
A
) is
valid for all the images of the row.
The PA technique demonstrates a good capability in revealing charcoal traits hidden
under the gypsum coats (Figure 3C). In the top raw (Figure 3), a 80-micron thick gypsum
layer covers both charcoal and sinopia lines. A charcoal line hidden under one and two
layers of gypsum cover (respectively 80 and 190
µ
m thick) is detectable by the PA technique,
as shown in Figure 3C. This method presents a good capability in revealing charcoal under
the gypsum coats and even allows clear detectability of charcoal in lower concentration
(mixed with sinopia) under two layers of gypsum (thickness 190
µ
m). At the same time,
sinopia is not detectable by photoacoustic imaging at the exploited wavelength (1064 nm)
but is detectable by near-infrared reflectance imaging. The combined PA imaging and
NIR reflectance imaging approach is shown to be helpful here in the detection of both
underdrawing materials (charcoal and sinopia) and making a hypothesis about its type.
Visible images of Sample 3 (Figure 4A–C) show yet another simulated scenario in
which the underdrawing traits, made with sinopia and with a charcoal/sinopia mixture,
are covered first with an Egyptian blue fresco paint (Figure 4B) and then with 80-micron

J. Imaging 2021,7, 250 7 of 12
thick gypsum layer (Figure 4C). Both PA and NIR images were processed in equal way by
means of ImageJ, applying 1.0 contrast stretching with min value: 10.826. The PA image
of the stratified sample, Figure 4D, reveals the mixed charcoal/sinopia trait whereas the
sinopia lines are visible in NIR reflectograms (Figure 4E,F).
J. Imaging 2021, 7, x FOR PEER REVIEW 7 of 12
layers of gypsum cover (respectively 80 and 190 μm thick) is detectable by the PA tech-
nique, as shown in Figure 3C. This method presents a good capability in revealing char-
coal under the gypsum coats and even allows clear detectability of charcoal in lower con-
centration (mixed with sinopia) under two layers of gypsum (thickness 190 μm). At the
same time, sinopia is not detectable by photoacoustic imaging at the exploited wavelength
(1064 nm) but is detectable by near-infrared reflectance imaging. The combined PA imag-
ing and NIR reflectance imaging approach is shown to be helpful here in the detection of
both underdrawing materials (charcoal and sinopia) and making a hypothesis about its
type.
Figure 3. Revealing hidden sketch drawings under gypsum coat in Sample 2: (A) visible image prior to gypsum applica-
tion; (B) visible image after gypsum coat application; (C) PA imaging result; (D,E) NIR reflectance images. The scale bar
in (A) is valid for all the images of the row.
Visible images of Sample 3 (Figure 4A–C) show yet another simulated scenario in
which the underdrawing traits, made with sinopia and with a charcoal/sinopia mixture,
are covered first with an Egyptian blue fresco paint (Figure 4B) and then with 80-micron
thick gypsum layer (Figure 4C). Both PA and NIR images were processed in equal way
by means of ImageJ, applying 1.0 contrast stretching with min value: 10.826. The PA image
of the stratified sample, Figure 4D, reveals the mixed charcoal/sinopia trait whereas the
sinopia lines are visible in NIR reflectograms (Figure 4E,F).
Figure 4. Revealing hidden sketch drawings under fresco paint and gypsum coat in Sample 3: visible images of (A) initial
drawing on the substrate; (B) after fresco paint application; (C) after gypsum coat application; (D) PA image of (C); (E,F)
NIR images of (C). The scale bar in (A) is valid for all the images.
Figure 4.
Revealing hidden sketch drawings under fresco paint and gypsum coat in Sample 3: visible images of (
A
) initial
drawing on the substrate; (
B
) after fresco paint application; (
C
) after gypsum coat application; (
D
) PA image of (
C
);
(E,F) NIR
images of (C). The scale bar in (A) is valid for all the images.
3.3. Imaging Charcoal and Sinopia Drawings under Limewash Coat: Samples 4 and 5
The next step of our research was aimed at revealing the underdrawings under the
limewash coat. Such simulation is represented in Sample 4, in which sinopia and charcoal
traits are concealed by a 60-microns thick limewash coat (Figure 5A,B top raw) and mixed
sinopia/charcoal under a 140 microns limewash (Figure 5A,B bottom raw). Both PA and
NIR images were processed in equal way by the means of ImageJ, applying 1.0% contrast
stretching and 0.9 gamma correction.
J. Imaging 2021, 7, x FOR PEER REVIEW 8 of 12
3.3. Imaging Charcoal and Sinopia Drawings under Limewash Coat: Samples 4 and 5
The next step of our research was aimed at revealing the underdrawings under the
limewash coat. Such simulation is represented in Sample 4, in which sinopia and charcoal
traits are concealed by a 60-microns thick limewash coat (Figure 5A,B top raw) and mixed
sinopia/charcoal under a 140 microns limewash (Figure 5A,B bottom raw). Both PA and
NIR images were processed in equal way by the means of ImageJ, applying 1.0% contrast
stretching and 0.9 gamma correction.
Figure 5. Revealing hidden sketch drawings under limewash coat in Sample 4: (A) visible image prior to limewash appli-
cation; (B) visible image after limewash coat application; (C) PA imaging result; (D,E) NIR reflectance images. The scale
bar in (A) is valid for all the images of the row.
PA imaging reveals the charcoal underdrawing in both scenarios, also under two
layers of gypsum, as displayed in Figure 5C. Sinopia underdrawing is detectable only by
NIR reflectance imaging at lower wavelengths, at 1050 nm (Figure 5D) and best defined
at 850 nm (Figure 5E). It is well known that depending on its concentration, sinopia be-
comes progressively transparent as a function of wavelength.
For the abovementioned reason, we simulated stratified Sample 5, where two con-
cealing layers (raw sienna fresco paint and 70-micron thick limewash coat) hide the char-
coal traits (Figure 6A–C). Indeed, in such a sample, both PA (Figure 6D) and NIR (Figure
6E,F) imaging are capable of revealing the charcoal sketches. We present here the PA and
NIR images processed in equal way by the means of ImageJ, applying 1.0% contrast
stretching. The charcoal line edges appear slightly blurred which is probably attributable
to the fresco execution method causing the dispersion of black pigment particles, as best
revealed by PA image in Figure 6D and by NIR image at 1830 nm in Figure 6F. The best
NIR result is observed at 1830 nm (Figure 6F), where limewash is known to show the
highest transparency [20].
Figure 5.
Revealing hidden sketch drawings under limewash coat in Sample 4: (
A
) visible image prior to limewash
application; (
B
) visible image after limewash coat application; (
C
) PA imaging result; (
D
,
E
) NIR reflectance images. The
scale bar in (A) is valid for all the images of the row.
PA imaging reveals the charcoal underdrawing in both scenarios, also under two
layers of gypsum, as displayed in Figure 5C. Sinopia underdrawing is detectable only by
NIR reflectance imaging at lower wavelengths, at 1050 nm (Figure 5D) and best defined at

J. Imaging 2021,7, 250 8 of 12
850 nm (Figure 5E). It is well known that depending on its concentration, sinopia becomes
progressively transparent as a function of wavelength.
For the abovementioned reason, we simulated stratified Sample 5, where two conceal-
ing layers (raw sienna fresco paint and 70-micron thick limewash coat) hide the charcoal
traits (Figure 6A–C). Indeed, in such a sample, both PA (Figure 6D) and NIR (
Figure 6E,F
)
imaging are capable of revealing the charcoal sketches. We present here the PA and NIR
images processed in equal way by the means of ImageJ, applying 1.0% contrast stretch-
ing. The charcoal line edges appear slightly blurred which is probably attributable to
the fresco execution method causing the dispersion of black pigment particles, as best
revealed by PA image in Figure 6D and by NIR image at 1830 nm in Figure 6F. The best NIR
result is observed at 1830 nm (Figure 6F), where limewash is known to show the highest
transparency [20].
J. Imaging 2021, 7, x FOR PEER REVIEW 9 of 12
Figure 6. Revealing hidden sketch drawings under limewash coat in Sample 5: visible images of (A) initial charcoal sketch
drawing; (B) after fresco paint application; (B) after limewash coat application; (C) PA image; (D,E) NIR reflectance images.
The scale bar in (A) is valid for all the images of the row. The red arrows in (D,F) indicate the blurred edges of the under-
drawing, which may be the result of dispersion of charcoal particles when applying by brush the fresco layer.
4. Discussion of the Results
We applied the novel epi-illumination PA imaging technique to ad-hoc designed
wall painting mock-ups with concealed underdrawings and fresco layers: overpainted,
covered by gypsum and limewash coats. The method proved powerful in revealing hid-
den sketch materials characterized by strong absorption properties at the working excita-
tion wavelength of the instrument (1064 nm). Therefore, detection of graphite underdraw-
ing is feasible with both PA and NIR techniques, whereas sinopia is detectable only by
NIR at lower wavelengths (the best result at 850–950 nm), where this material shows
stronger absorption properties. NIR imaging proved particularly useful in revealing both
characteristic traditional fresco underdrawing materials (charcoal and sinopia) under thin
hiding coats (<80 μm). This can be explained by the fact that in the near-infrared range,
thin gypsum and lime layers scatter and absorb less light, which results in their increased
transparency (when compared to the visible range). At the wall surface, however, the
near-infrared radiation is reflected from the ground and absorbed by the underdrawing.In
NIR reflectography, revelation of hidden drawings, located at the wall ground, is subject
to both near-infrared transparency of hiding coats (due to low scattering and low absorb-
ance) and to the differences in near-infrared reflection/absorption properties of the wall
painting materials. While the performance of NIR imaging is subject to optical properties
of the hiding and hidden features, the performance of PA imaging is based on the optical
and acoustic properties of the materials. When increasing the coat thickness (up to 190 in
gypsum and up to 140 in limewash coats), contrast enhancement both in PA and in NIR
images becomes helpful for better revelation of charcoal traits. Revealing materials with a
low absorption contrast at 1064 nm (e.g., sinopia) is expected to be addressed in future
development of the reflection mode PA method at more excitation wavelengths.
We further note that the PA imaging complements the traditional NIR imaging tech-
nique in studying the peculiar features of charcoal traits, e.g., definition of the line, blurred
edges, and spolvero dots. These details are helpful for the study of the execution technique,
author attribution other art historical and conservation queries.
The effectiveness of both techniques is determined by the combination of the physical
properties of constituent materials at the system operating wavelengths: optical proper-
ties for NIR imaging and both optical and ultrasonic properties for PA imaging. This dif-
ference can help understand divergences in the detectability of hidden features by the two
techniques even at same thickness of hiding material and similar working wavelengths
(e.g., between PA image at 1064 and NIR image at 1050 nm). We cannot exclude the fact
that the result might be also influenced by the grain size and morphology of top layers
[43], which is subject of further studies by authors.
This paper deals with the first proof-of-concept study and at the initial step, the use
of water as an immersion medium, was preferred due to the simplicity of handling and
Figure 6.
Revealing hidden sketch drawings under limewash coat in Sample 5: visible images of (
A
) initial charcoal
sketch drawing; (
B
) after fresco paint application; (
B
) after limewash coat application; (
C
) PA image; (
D
,
E
) NIR reflectance
images. The scale bar in (
A
) is valid for all the images of the row. The red arrows in (
D
,
F
) indicate the blurred edges of the
underdrawing, which may be the result of dispersion of charcoal particles when applying by brush the fresco layer.
4. Discussion of the Results
We applied the novel epi-illumination PA imaging technique to ad-hoc designed
wall painting mock-ups with concealed underdrawings and fresco layers: overpainted,
covered by gypsum and limewash coats. The method proved powerful in revealing hidden
sketch materials characterized by strong absorption properties at the working excitation
wavelength of the instrument (1064 nm). Therefore, detection of graphite underdrawing
is feasible with both PA and NIR techniques, whereas sinopia is detectable only by NIR
at lower wavelengths (the best result at 850–950 nm), where this material shows stronger
absorption properties. NIR imaging proved particularly useful in revealing both character-
istic traditional fresco underdrawing materials (charcoal and sinopia) under thin hiding
coats (<80
µ
m). This can be explained by the fact that in the near-infrared range, thin
gypsum and lime layers scatter and absorb less light, which results in their increased
transparency (when compared to the visible range). At the wall surface, however, the
near-infrared radiation is reflected from the ground and absorbed by the underdrawing. In
NIR reflectography, revelation of hidden drawings, located at the wall ground, is subject to
both near-infrared transparency of hiding coats (due to low scattering and low absorbance)
and to the differences in near-infrared reflection/absorption properties of the wall painting
materials. While the performance of NIR imaging is subject to optical properties of the
hiding and hidden features, the performance of PA imaging is based on the optical and
acoustic properties of the materials. When increasing the coat thickness (up to 190 in
gypsum and up to 140 in limewash coats), contrast enhancement both in PA and in NIR
images becomes helpful for better revelation of charcoal traits. Revealing materials with
a low absorption contrast at 1064 nm (e.g., sinopia) is expected to be addressed in future
development of the reflection mode PA method at more excitation wavelengths.
We further note that the PA imaging complements the traditional NIR imaging tech-
nique in studying the peculiar features of charcoal traits, e.g., definition of the line, blurred

J. Imaging 2021,7, 250 9 of 12
edges, and spolvero dots. These details are helpful for the study of the execution technique,
author attribution other art historical and conservation queries.
The effectiveness of both techniques is determined by the combination of the physical
properties of constituent materials at the system operating wavelengths: optical properties
for NIR imaging and both optical and ultrasonic properties for PA imaging. This differ-
ence can help understand divergences in the detectability of hidden features by the two
techniques even at same thickness of hiding material and similar working wavelengths
(
e.g., between
PA image at 1064 and NIR image at 1050 nm). We cannot exclude the fact
that the result might be also influenced by the grain size and morphology of top layers [
43
],
which is subject of further studies by authors.
This paper deals with the first proof-of-concept study and at the initial step, the use
of water as an immersion medium, was preferred due to the simplicity of handling and
due to its optimal bond performance for signal propagation. In the future, we expect
further development of the technique with the implementation of air-coupled transducers,
which have been already introduced in the transmission mode photoacoustic imaging
set up [
30
]. Moreover, the same detection technology of PA signals has been recently
employed for the investigation of restoration operations in a historical oil painting from
the 19th century, demonstrating the high potential of air-coupled detectors in transmission
geometries [
44
]. Similar implementations in reflection-mode may enable absolutely non-
contact and non-invasive PA imaging of painted artifacts with a relative compromise as
regards the detection sensitivity and the spatial resolution of the system.
5. Conclusions
Here, non-invasive PA imaging was tested for the first time tested as a tool for reveal-
ing hidden underdrawings in wall paintings with a complex stratigraphy. In this study, we
exploited the prototype in its new epi-illumination configuration. The first experimental
results presented here demonstrate the effectiveness of PA imaging in revealing hidden
underdrawings in simulated retouched (overpainted) and concealed (by gypsum and
limewash) wall paintings.
The non-invasive approach by the novel reflection-mode PA imaging technique proved
capable of revealing the simulated hidden features in fresco and secco wall paintings, pre-
pared following historical methodologies. The results were complemented and validated
by the well-established NIR imaging technique. PA and NIR imaging proved complemen-
tary here in revealing hidden sketches (graphite, charcoal, sinopia, mix of charcoal and
sinopia) under overpaint, gypsum and limewash coats, also in the presence of fresco paint
layers. Under thicker concealing layers of gypsum (up to 190
µ
m), PA proved efficient in
revealing charcoal traits (also in lower concentration, mixed with sinopia), showing a good
contrast also under thicker concealing layers of gypsum. The current limitations of the
novel reflectance photoacoustic imaging prototype, discussed above, are expected to be
addressed in future development of the method. On the basis of the obtained results, we
expect that this efficient non-invasive diagnostic tool will represent a breakthrough in Her-
itage Science for diagnostic applications to complex wall paintings, for their conservation
and, if reasonable, for their safe uncovering.
Author Contributions:
Conceptualization, A.C., G.J.T. and J.S.; methodology, A.C., G.J.T., E.K., R.F.
and J.S.; software, G.J.T.; validation, G.J.T., E.K., R.F. and J.S.; formal analysis, A.C., G.J.T. and E.K.;
investigation, A.C., G.J.T. and E.K.; resources, A.C., J.S.; data curation, A.C.; writing—original draft
preparation, A.C.; writing—review and editing, G.J.T., E.K., R.F., G.Z. and J.S.; visualization, A.C.
and G.J.T.; supervision, G.Z. and J.S.; project administration, R.F. and J.S.; funding acquisition, J.S. All
authors have read and agreed to the published version of the manuscript.

J. Imaging 2021,7, 250 10 of 12
Funding:
The research was supported by the H2020 Laserlab Europe [EC-GA 871124], project number
2663, by the funds of POR FSE 2014–2020 Giovanisìof the Tuscany Region (Italy) within the frames
of program “CNR4C” (Joint Advanced Education Project 249795); by the H2020 FETOPEN project
“Dynamic” (EC-GA-863203), the NSRF 2014–2020 “BIOIMAGING-GR” (MIS 5002755), HELLAS CH
(MIS 5002735), “INNOVAPROTECT” (MIS 5030524) funded by the Operational Program “Compet-
itiveness, Entrepreneurship, and Innovation” under the calls Reinforcement of the Research and
Innovation Infrastructure and RESEARCH–CREATE–INNOVATE, respectively, and co-financed by
Greece and the European Union (European Regional Development Fund).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement:
The data that support the findings of this study are available from the
corresponding author upon reasonable request.
Acknowledgments:
The authors would like to thank Diego Ivan Quintero Balbas, Mariarosa Lan-
franchi, Maria Cristina Gigli and Loris Nicoletti for their contribution in the realization of wall
painting mock-ups. Furthermore, the authors acknowledge Enrico Pampaloni who performed layer
thickness calculations from the set of microprofilometry data.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Cennini, C. The Craftman’s Handbook: II Libro dell’Arte’ Cennino a’Andrea Cennini; Thompson, D.V., Ed.; Dover: New York, NY, USA,
1960; ISBN 9780486200545.
2.
Wallert, A.; Hermens, E.; Peek, M. (Eds.) Historical Painting Techniques, Materials and Studio Practice, Preprints of
a Symposium
Held at the University of Leiden, The Netherlands, 26–29 June 1995; Getty Conservation Institute: Malibu, CA, USA, 1995;
ISBN 9780892363223.
3.
Basile, G. (Ed.) Giotto Nella Cappella Scrovegni: Materiali per la Tecnica Pittorica, Studi e Ricerche Dell’istituto Centrale per Il Restauro;
Istituto Poligrafico e Zecca dello Stato: Rome, Italy, 2005.
4. Botticelli, G. Metodologia e Restauro Delle Pitture Murali; Edizioni Centro Di: Firenze, Italy, 1992.
5.
Piovesan, R.; Mazzoli, C.; Maritan, L.; Cornale, P. Fresco and Lime Paint: An Experimental Study and Objective Criteria for
Distinguishing Between These Painting Techniques. Archaeometry 2012,54, 723–736. [CrossRef]
6.
Rattazzi, D.; Maisano, D. Quando Il Muro Nasconde L’affresco; Informazioni, Commenti, Inchieste sui Beni Culturali XII, 4; IBC:
Bologna, Italy, 2004.
7.
Rogoz, J.; Cupa, A.; Wiecek, B. The Analysis of Mural Painting “The Crucifiction” in St. John the Baptist and St. John the
Evangelist Basilica in Toru ´n; QIRT2008 Proceedings. 2008. Available online: http://qirt.gel.ulaval.ca/archives/qirt2008/papers/
16_07_07.pdf (accessed on 1 October 2021).
8.
Smith, C. Cleaning Methods for the Removal of Limewash from Painted Plaster Surfaces: Utilizing Ion Exchange Resins on the
Interior Architectural Finishes of the Capilla de Nuestra Señora del Rosario in Iglesia San José, in San Juan, Puerto Rico. Master’s
Thesis, University of Pennsylvania, Philadelphia, PA, USA, 2009.
9.
Walsh, M. Emergency Conservation of Medieval Murals in Northern Cyprus: St. Anne’s Church, Famagusta. In Reuso 2015,
Proceedings of the 3rd International Conference on Documentation, Conservation and Restoration of the Architectural Heritage and Landscape
Protectionm, Valencia, Spain, 22–24 October 2015; Editorial Universitat Politècnica de València: Valencia, Spain, 2015.
10.
Davin, T.; Serio, B.; Guida, G.; Pina, V. Spatial Resolution Optimization of a Cooling-Down Thermal Imaging Method to Reveal
Hidden Academic Frescoes. Int. J. Therm. Sci. 2017,112, 188–198. [CrossRef]
11.
Baggio, L. Iconografia di sant’Antonio al Santo di Padova nel XIII e XIV Secolo: Spazi, Funzioni, Messaggi Figurati, Committenze.
Ph.D. Thesis, Universitàdi Padova, Padua, Italy, 2014.
12.
Poldi, G.; Villa, G.C.F. Dalla Conservazione Alla Storia Dell’arte. Riflettografia e Analisi Noninvasive per lo Studio Dei Dipinti; Edizioni
della Normale: Pisa, Italy, 2006; ISBN 8876422056.
13.
Vallet, J.-M.; Detalle, V.; De Luca, L.; Bodnar, I.-L.; Guillon, O.; Trichereau, B.; Mouhoubi, K.; Beaumont, M.N.; Syvilay, D.;
Giovannacci, D.; et al. Development of a NDT toolbox dedicated to the conservation of wall paintings: Application to the frescoes
chapel in the Charterhouse of Villeneuve-lez-Avignon (France). Digit. Herit. 2013,2, 67–74.
14.
Striova, J.; Dal Fovo, A.; Fontana, R. Reflectance imaging spectroscopy in heritage science. Riv. Nuovo Cim.
2020
,
43, 515–566. [CrossRef]
15.
Liang, H. Advances in multispectral and hyperspectral imaging for archaeology and art conservation. Appl. Phys. A
2012
,
106, 309–323. [CrossRef]

J. Imaging 2021,7, 250 11 of 12
16.
Daffara, C.; Pampaloni, E.; Pezzati, L.; Barucci, M.; Fontana, R. Scanning multispectral IR reflectography SMIRR: An advanced
tool for art diagnostics. Acc. Chem. Res. 2010,43, 847–856. [CrossRef] [PubMed]
17.
Legnaioli, S.; Lorenzetti, G.; Cavalcanti, G.H.; Grifoni, E.; Marras, L.; Tonazzini, A.; Salerno, E.; Pallecchi, P.; Giachi, G.;
Palleschi, V.
Recovery of archaeological wall paintings using novel multispectral imaging approaches. Herit. Sci. 2013,1, 33. [CrossRef]
18.
Chaban, A.; Deiana, R.; Parisatto, M.; Asscher, Y. Analysis of on-site multispectral images: A case study of degraded wall paintings
in Sarno Baths, Pompeii. In Acta Artis Academica, Proceedings of the Conference ALMA 2017 “Painting as a Story”, Brno, Czech Republic,
31 May–3 June 2017; Academy of Fine Arts in Prague: Prague, Czech Republic, 2017; pp. 151–160, ISBN 978-80-87108-75-8.
19.
Ricci, M.; Laureti, S.; Malekmohammadi, H.; Sfarra, S.; Lanteri, L.; Colantonio, C.; Calabrò, G.; Pelosi, C. Surface and Interface
Investigation of a 15th Century Wall Painting Using Multispectral Imaging and Pulse-Compression Infrared Thermography.
Coatings 2021,11, 546. [CrossRef]
20.
Fontana, R.; Striova, J.; Barucci, M.; Pampaloni, E.; Raffaelli, M.; Pezzati, M.; Mariotti, P. Limewashed Mural Paintings as Seen by
VIS-IR Reflectography. 2015, Volume 9527, p. 95270Y. Available online: https://spie.org/Publications/Proceedings/Paper/10.1
117/12.2197995?SSO=1 (accessed on 1 October 2021).
21.
Koch Dandolo, C.; Jepsen Uhd, P. Wall Painting Investigation by Means of Non-invasive Terahertz Time-Domain Imaging
(THz-TDI): Inspection of Subsurface Structures Buried in Historical Plasters. Appl. Opt. 2015,54, 5123–5129. [CrossRef]
22.
Giovannacci, D.; Cheung, H.C.; Walker, G.C.; Bowen, J.W.; Martos-Levif, D.; Brissaud, D.; Cristofol, L.; Mélinge, Y. Time-domain
imaging system in the terahertz range for immovable cultural heritage materials. Strain 2019,55, e12292. [CrossRef]
23.
Filippidis, G.; Massaouti, M.; Selimis, A.; Gualda, E.J.; Manceau, J.-M.; Tzortzakis, S. Nonlinear imaging and THz diagnostic tools
in the service of Cultural Heritage. Appl. Phys. A 2012,106, 257–263. [CrossRef]
24.
Redo-Sanchez, A.; Heshmat, B.; Aghasi, A.; Naqvi, S.; Zhang, M.; Romberg, J.; Raskar, R. Terahertz time-gated spectral imaging
for content extraction through layered structures. Nat. Commun. 2016,7, 1266. [CrossRef] [PubMed]
25.
Liang, H.; Lange, R.; Howard, H.; Spooner, J. Non-Invasive Investigations of a Wall Painting Using Optical Coherence Tomography
and Hyperspectral Imaging. 2011, p. 80840F. Available online: https://www.spiedigitallibrary.org/conference-proceedings-of-
spie/8084/80840F/Non-invasive-investigations-of-a-wall-painting-using-optical-coherence/10.1117/12.890088.short?SSO=1
(accessed on 1 October 2021).
26.
Liang, H.; Cid, M.G.; Dobre, G.M.; Podoleanu, A.G.; Pedro, J.; Saunders, D. En-face optical coherence tomography—A novel
application of non-invasive imaging to art conservation. Opt. Express 2005,13, 6133–6144. [CrossRef]
27.
Targowski, P.; Iwanicka, M. Optical coherence tomography: Its role in the non-invasive structural examination and conservation
of cultural heritage objects–A review. Appl. Phys. A 2012,106, 265–277. [CrossRef]
28.
Tserevelakis, G.J.; Chaban, A.; Klironomou, E.; Melessanaki, K.; Striova, J.; Zacharakis, G. Revealing Hidden Features in Multilay-
ered Artworks by Means of an Epi-Illumination Photoacoustic Imaging System. J. Imaging 2021,7, 183. [CrossRef] [PubMed]
29.
Tserevelakis, G.J.; Vrouvaki, I.; Siozos, P.; Melessanaki, K.; Hatzigiannakis, K.; Fotakis, C.; Zacharakis, G. Photoacoustic imaging
reveals hidden underdrawings in paintings. Sci. Rep. 2017,7, 747. [CrossRef]
30.
Tserevelakis, G.J.; Siozos, P.; Papanikolaou, A.; Melessanaki, K.; Zacharakis, G. Non-invasive photoacoustic detection of hidden
underdrawings in paintings using air-coupled transducers. Ultrasonics 2019,98, 94–98. [CrossRef]
31.
Tserevelakis, G.J.; Tsagkaraki, M.; Siozos, P.; Zacharakis, G. Uncovering the hidden content of layered documents by means of
photoacoustic imaging. Strain 2019,55, e12289. [CrossRef]
32.
Dal Fovo, A.; Tserevelakis, G.J.; Papanikolaou, A.; Zacharakis, G.; Fontana, R. Combined photoacoustic imaging to delineate the
internal structure of paintings. Opt. Lett. 2019,44, 919–922. [CrossRef] [PubMed]
33.
Tserevelakis, G.J.; Dal Fovo, A.; Melessanaki, K.; Fontana, R.; Zacharakis, G. Photoacoustic signal attenuation analysis for the
assessment of thin layers thickness in paintings. J. Appl. Phys. 2018,123, 123102. [CrossRef]
34.
Shankar, H.; Pagel, P.S.; Warner, D.S. Potential Adverse Ultrasound-related Biological Effects: A Critical Review. Anesthesiology
2011,115, 1109–1124. [CrossRef] [PubMed]
35.
Hosseinaee, Z.; Le, M.; Bell, K.; Haji Reza, P. Towards Non-Contact Photoacoustic Imaging. Photoacoustics
2020
,
20, 100207. [CrossRef] [PubMed]
36.
Striova, J.; Pezzati, L.; Pampaloni, E.; Fontana, R. Synchronized Hardware-Registered VIS-NIR Imaging Spectroscopy and 3D
Sensing on a Fresco by Botticelli. Sensors 2021,21, 1287. [CrossRef] [PubMed]
37.
Daffara, C.; Fontana, R. Multispectral Infrared Reflectography to Differentiate Features in Paintings. Microsc. Microanal.
2011
,
17, 691–695. [CrossRef]
38.
Striova, J.; Fontana, R.; Barucci, M.; Felici, A.; Marconi, E.; Pampaloni, E.; Raffaelli, M.; Riminesi, C. Optical devices provide
unprecedented insights into the laser cleaning of calcium oxalate layers. Microchem. J. 2016,124, 331–337. [CrossRef]
39.
Daffara, C.; Fontana, R.; Di Crescenzo, M.M.; Scrascia, S.; Zendri, E. Optical Techniques for the Characterization of Surface-
Subsurface Defects in Painting Layers; Proc. SPIE 8084. 2011. Available online: https://www.spiedigitallibrary.org/conference-
proceedings-of-spie/8084/1/Optical-techniques-for-the-characterization- of-surface- subsurface-defects- in/10.1117/12.889959.
short (accessed on 1 October 2021).
40.
Bevilacqua, N.; Borgioli, L.; Adrover Garcia, I. I Pigmenti Nell’arte: Dalla Preistoria Alla Rivoluzione Industriale; II Prato: Saonara,
Italy, 2010; 302p, ISBN 8863360901.
41. Djuriši´c, A.B.; Li, E.H. Optical properties of graphite. J. Appl. Phys. 1999,85, 7404–7410. [CrossRef]

J. Imaging 2021,7, 250 12 of 12
42.
Hagemann, H.J.; Gudat, W.; Kunz, C. Optical Constants from the Far Infrared to the X-ray Region: Mg, Al, Cu, Ag, Au, Bi, C, and
Al2O3.J. Opt. Soc. Am. 1975,65, 742–744. [CrossRef]
43.
Agudo, R.; lvarez-Lloret Ibañ ez-Velasco, A.; Ortega-Huertas, M. Crystallographic Control in the Replacement of Calcite by
Calcium Sulfate. Crystal 2016,16, 4950–4959.
44.
Dal Fovo, A.; Tserevelakis, G.J.; Klironomou, E.; Zacharakis, G.; Fontana, R. First combined application of photoacoustic and
optical techniques to the study of an historical oil painting. Eur. Phys. J. Plus 2021,136, 757. [CrossRef]