Imma Perfetto
Cosmos science journalist
A PhD student in the US has turned an art restoration hobby into a new method to repair damaged paintings faster and more cheaply than conventional approaches.
The method involves using artificial intelligence to create a “digitally restored” version of the damaged painting, which is then printed onto a thin polymer film and applied directly on top of the art like a “mask” to correct areas of paint loss or discolouration.
The approach could allow conservators to restore lower value paintings that would otherwise remain behind closed doors due to limited conservation budgets, which prioritise more famous, valuable artworks.
Alex Kachkine, graduate student in mechanical engineering at Massachusetts Institute of Technology (MIT), says: “Hopefully with this new method, there’s a chance we’ll see more art, which I would be delighted by.”
The method is presented in a paper in the journal Nature.
Over the decades and centuries following their creation, paintings are exposed to light, temperature fluctuations, varying air quality, and chemical reactions in the paint layers, which causes cracking, flaking, and discolouration.
“Damaged paintings can be restored by retouching – a combination of infilling to replace missing areas of paint, and painting over discolouration,” Associate Professor Hartmut Kutzke of the Museum of Cultural History at the University of Oslo, Norway wrote in a related Nature News & Views article.
“Restoring paintings manually is an expensive and labour-intensive process – and, if done badly, it can damage the piece of art permanently.”
Kachkine restores paintings as a hobby using these traditional hand-painting techniques – a process that can take months to years of work.
Researchers have developed generative AI programs to create digitally restored versions of damaged paintings before. But, while these tools provide a vision of what a finished restoration could look like, they do not remove the need for slow, meticulous retouching.
Kachkine’s new method does. In the new study, the digital restoration was physically applied onto an original painting – a highly damaged oil-on-panel attributed to the Master of the Prado Adoration from the late 15th century.
“This painting is almost 600 years old and has gone through conservation many times,” he says.
“In this case there was a fair amount of overpainting, all of which has to be cleaned off to see what’s actually there to begin with.”
Kachkine scanned the cleaned painting and used existing AI algorithms to analyse it, creating a virtual version of what the painting likely looked like in its original state. He then developed a software that created a map of regions on the original painting that were lost and required infilling, as well as the exact colours needed to match the digitally restored version.
The method identified 5,612 separate regions and filled them using 57,314 different colours. This map was then translated into a 2-layer mask which was printed onto thin polymer-based films using commercial inkjets.
The printed films were made from materials that can be dissolved with conservation-grade solutions and were adhered to the painting with conservation-grade varnish. Crucially, this means the intervention can be reversed by future conservationists who may wish to reveal the original, damaged work underneath.
“Because there’s a digital record of what mask was used, in 100 years, the next time someone is working with this, they’ll have an extremely clear understanding of what was done to the painting,” says Kachkine.
“And that’s never really been possible in conservation before.”
The entire process took 3.5 hours, which he estimates is about 66 times faster than traditional restoration methods.
In his related article, Kutzke writes that further work will be needed to understand the effect of the films on the painting itself.
“To prevent the printed ink from smearing, the mask was sprayed with conservation-grade varnish, so there was no direct interaction between the paint and the mask,” he writes.
“However, when a painting is sealed in its frame, the layer of air between the painting and the mask forms a microclimate. Humidity and slowly evaporating chemical compounds in this trapped layer of air could result in slow damage to the painting.
“Artworks restored by this technique should thus be monitored carefully.”
Also, this approach would only be suitable for similarly flat oil paintings created on solid bases.
“Topographically complex oil paintings, such as impressionist works with extensive use of impasto, also cannot be masked … without optical defects stemming from regional misalignment and poor surface adhesion,” Kachkine writes in his research.
“Mask removal for these works could also endanger fragile textural elements.”
Kachkine emphasises that any application of his new method should also be done in consultation with conservators with knowledge of a painting’s history and origins to ensure that the final work is in keeping with an artist’s style and intent.
“It will take a lot of deliberation about the ethical challenges involved at every stage in this process to see how this can be applied in a way that’s most consistent with conservation principles,” he says.
“We’re setting up a framework for developing further methods. As others work on this, we’ll end up with methods that are more precise.”
