A preview of this full-text is provided by Optica Publishing Group.
Content available from Applied Optics
This content is subject to copyright. Terms and conditions apply.
Research Article Vol. 61, No. 8 / 10 March 2022 / Applied Optics E33
Pulsed thermography and infrared reflectography:
comparative results for underdrawing
visualization in paintings
Jacopo Melada,1Marco Gargano,2,*AND Nicola Ludwig2
1Dipartimento di Scienze della Terra “Ardito Desio,” Università degli Studi di Milano, Milano, Italy
2Dipartimento di Fisica “Aldo Pontremoli,” Università degli Studi di Milano, Milano, Italy
*Corresponding author: marco.gargano@unimi.it
Received 6 October 2021; revised 26 December 2021; accepted 3 January 2022; posted 6 January 2022; published 15 February 2022
Reflectographic analyses applied on paintings can be performed using cameras equipped with different
detectors with different abilities in detecting and visualizing underdrawings, repainting, restorations, and other
nonvisible information. In this research, the results obtained through thermographic imaging followed by
statistical imaging postprocessing methods have been compared with those obtained with traditional
reflectographic methods in the short-wave infrared range. The comparison has been performed studying the
thermal sequence after a single pulse of light with a different spectrum of ad hoc mock-ups. Results showed that
for limited cases, signal-to-noise ratio seems to be more relevant in obtaining reliable images of underdrawings
with respect to the effect of optical absorption of visible light by painting layers. © 2022 Optica Publishing Group
https://doi.org/10.1364/AO.445047
1. INTRODUCTION
Scientific methods used in the examination of easel paintings
have evolved over the last 50 years toward a totally noninvasive
way of inspection by introducing ever more devoted techniques
for chemical analyses of pigments and binders thanks to mul-
tispectral imaging [1–4]. In reflectography, the InGaAs-based
detectors, sensitive in short-wave infrared (SWIR, 0.78–
1.70 µm), have established themselves as the most effective
in visualizing important information hidden by most of the
pigment used in oil pictorial layers [5–8]. SWIR extended
sensibility up to 2.5 µm can be achieved using other semicon-
ductor detectors such as PtSi or germanium [9], and this value
become an explicit limit of investigation caused by glass lens
transmittance, absorbance by binders, and the noise due to
thermal emission given by the blackbody at room temperature.
This has limited the attempt to extend in the middle IR band
for the underdrawings’ visualization by means of a short-wave
thermal camera with passive technique [10]. In this context, in
the last 30 years, most efforts have been devoted to improving
both spatial resolution of IR images [11] and portability of
SWIR instrumentation [12] for the application on wall paint-
ings. In this way, the techniques have become more and more
routinely and essentially oriented to the needs of scholars and
art historians for the best underdrawing detection and repre-
sentation. The improvements of the detectors in mid-wave
infrared (MWIR) and long-wave infrared (LWIR) coupled with
fast digital postprocessing using infrared thermography (IRT)
on artworks [13–15] have eventually changed in recent years,
giving IR inspection of paintings new chances of application in
conservation projects. Relevant works have been proposed to
the scientific community considering the issues of long-distance
inspection [4,12,16] as well as pictorial film detachments [15].
Finally, the use of statistical analysis of thermal image sequences
after short thermal pulse stimulation has shown to be helpful in
both diagnosis of pictorial layers and underdrawings detection
when traditional SWIR reflectography fails [16–18]. In a recent
research work that addresses a combination of IRT and multi-
spectral imaging, a short review of tools and methods used in IR
imaging for painting inspection can be found [19]. In this work,
we present quantitative and qualitative comparison between
traditional InGaAs high-resolution detectors and pulsed LWIR
thermography with different postprocessing algorithms applied
on different mock-ups prepared simulating different pictorial
layers and drawing materials.
2. MATERIALS
The mock-ups (R1, R2, R3) used in this laboratory test have
been specifically designed and realized in cooperation with
restorers in order to simulate 15th- through 17th-century paint-
ings for what concerns the thickness, the chemical composition,
the binders, and wood panel priming (imprimitura). As detailed
in Fig. 1, seven different drawing media have been applied above
three types of imprimitura: R1 with lead white in linseed oil
binder applied on a preparation of gesso and animal glue; R2
with 1:1 mixture of lead-white and carmine in linseed oil binder
1559-128X/22/080E33-06 Journal © 2022 Optica Publishing Group
