|
Hyper-Spectral Imaging (HSI) has emerged in the last decade as one of the most promising technologies for
diagnostics and documentation of polychrome surfaces. Despite the fact that presently HSI is a well-established
technique for non-invasive investigations on paintings, a number of technological issues remain open and are
still topics for on-going studies. In particular, it is known that high spatial resolution is a crucial parameter for
obtaining high quality images, whereas the possibility to identify pictorial materials strictly depends on the
spectral resolution and on the extent of the spectral region investigated. At the same time, by increasing the
sampling rates in both the spatial and spectral dimensions, the size of the data-set will be enlarged and the
acquisition times will be lengthened. As a consequence, a good compromise between the acquisition of highquality
data and their application should always be reached, taking into account the specific purposes of the HSI
application. The above questions are discussed in the present work, which illustrates two applications of the
latest version of a hyperspectral scanner designed at IFAC-CNR for the digitization of artworks. The prototype
has recently been upgraded, with new visualization software as well as mechanical and optical improvements.
This high performance system operates in the 400-1000nm spectral range, with a spectral resolution of about 2-3
nm and a spatial sampling of 0.1 mm over areas of about 1 m2. Three case-studies are presented, which highlight
the importance of both high spatial and high spectral sampling rate in hyperspectral imaging. Two of the
examples reported focus on the full exploitation of the spatial resolution: the first one is a study performed on a
small painting, dated from the eighteenth century and belonging to the Uffizi Gallery in Florence; the second
case-study refers to the valuable "Carrand diptych" (14th century) from the Bargello Museum in Florence. The
last application, instead, shows the crucial importance of a high spectral resolution to identify selected pigments
in the oil-painting "Ritratto di Maffeo Barberini", dated around 1596-1600, which has recently been attributed to
Caravaggio.
|
