Photography is a
recent form of art, but probably the most diffused in the world. Since its introduction it
has supplied us with a huge amount of documents of high cultural value. The ubiquitous
fruition of such documents requires as a first step the conversion of old printed material
into digital form, for successive manipulation and data management. Unfortunately, the
oldest photographic prints are based on fragile materials, which are affected by bad
environmental conditions. For this reason, in these years, the care for the preservation
and conservation of prints has increased. However there are many damaged pictures
that still need to be restored. With classical, physical restoration, the images are
dusted and cleaned, the gaps are filled up, the emulsion is consolidated, the gelatin is
fixed, and some parts are repainted by hand. This pipeline produces cleaner images, but is
extremely expensive, and often the damage is still visible.
The diffusion of
scanners and of software to manipulate the image opens a new way to the recovery of
photos. The old image, after the digitization, can be virtually restored and, if
necessary, reprinted. For this type of digital restoration, the goal is that a person
viewing only the restored version should not be able to realize where the changes are.
Notice that this case is different from the one of the physical restoration of originals,
where the original artwork needs to be preserved. Commercial software exists, which
proposes a heavy user-guided restoration, where the defects are not automatically traced,
and also all the corrections must be user-suggested. Virtual restoration becomes complex
and expensive, and it can be performed only by skilled personnel. Automated restoration is
hence required to obtain quick, simple and effective results.
The research
project SiRAD which is being supported by the Regione Friuli Venezia Giulia, in
cooperation with Fratelli Alinari SpA, and is entirely devoted to the study of novel
methods for the automated digital restoration of photographic archives. In this web page
we propose a dissertation about defect detection and correction in photographic prints. It
should be emphasized that at least three different advantages stem from the digital
restoration of photographic prints:
·
ubiquitous fruition is made easier, since a clean image typically
can be much more efficiently coded and transmitted or stored;
·
the performances of content description methods for the purpose
of retrieval and browsing are improved;
·
the visual quality of the output is improved, in particular on
low-cost portable displays.