If one had to choose a word to describe the work by Ingrid Daubechies, it could well be “interdisciplinarity”. This professor at Duke University (USA), who is currently the Chair of the International Mathematical Union (IMU), started her career as a theoretical physicist and then gradually moved on to mathematics, motivated by the pressing need for new tools of this type in her original discipline. In 2012, she was awarded the FBBVA Frontiers of Knowledge Prize in recognition of her work on wavelets, which has been applied for instance to the JPEG 2000 image compression standard.
“Image analysis can be used to distinguish an artist’s line or brushstroke”
Daubechies is also attracted by problems arising in other areas, such as art. “Somebody drew my attention to the fact that image analysis can be used to distinguish an artist’s line or brushstroke” in order to determine, for example, the authenticity of a work.
The latest work she has completed has to do with just that, and she will talk about it in her appearance at the AIMS 2014 Conference. Together with other collaborators from the University of Brussels (Belgium) and the North Carolina Museum of Art (USA), Daubechies has developed an algorithm that enables the original brushstrokes of artists to be seen by means of X-rays, which also reveal the techniques the artists used, the way in which the paints were prepared, and the state of conservation of the artworks in question.
Between the 12th and 17th centuries, European artists painted on wooden panels. Later, in the 19th and 20th centuries, conservators planed away these panels to make them as thin as possible and backed them with a latticework or stretcher that was inseparable from the work itself and protected it from damage. However, this type of backing makes it difficult to study the original work with X-rays, a technique much used at present to study the state of a painting.
Until now it was possible to eliminate the image of the latticework that showed up on the X-ray by hand, although this was a complicated procedure that could only be carried out on a limited number of paintings. Researchers thought that a more automated way of removing this image would be of great help to conservators and museum curators. The results obtained have turned out to be satisfactory and similar to the techniques employed to date, which shows that mathematics and art go hand in hand on more occasions than one might have imagined.
This text is an excerpt from an original article by Lorena Cabeza, ICMAT. Find the full text here.
Instituto de Ciencias Matemáticas (ICMAT)
Mathematical research center set up by the Consejo Superior de Investigaciones Científicas (CSIC – Higher Council for Scientific Research)
and three universities in Madrid: Autónoma University (UAM), Carlos III University (UC3M)
and Complutense University (UCM).