Algorithms Offer Glimpses Behind Artistic Styles

January 21, 2010

Researchers claim to have come up with mathematical algorithms that offer insights into the artistic styles of works of art. Their algorithms analyze so-called low-level pictorial information such as brush-stroke thicknesses, types of canvases, and compositions of palettes of colors. 

The idea behind this is to enable scholars to classify and research collections in museums, and to better understand relationships among people, computers, and art.

In 1933, mathematician George D. Birkhoff made a similar attempt by trying to formalize the notion of beauty, based on relationships between order and complexity. Max Bense made use of Birkhoff's work to try to come up with a measurement of information based on entropy (disorder or diversity).

Still out of reach, however, are algorithms that can analyze medium-level information that differentiates objects and scenes in pictures, as well as algorithms that distinguish among types of paintings (landscapes, portraits, still life, etc.). High-level information that takes into account historical contexts or has knowledge of artists' backgrounds and artistic developments is also elusive at the moment.

"It will never be possible to precisely determine mathematically an artistic period nor to measure the human response to a work of art, but we can look for trends," said Miquel Feixas (University of Girona).

Feixas and colleagues from Girona and from Germany's Max Planck Institute reported their findings in "Categorizing Art: Comparing Humans and Computers" (Computers & Graphics, issue 4, 2009).

Source: Science Codex (December 23, 2009).

Image via Wikipedia.