Sometimes it’s hard to find some truly interesting technology news. But here’s a recent story that certainly is: Serbian computer scientists have developed a computer algorithm to spot the difference between an original work of art and a copy of the same piece – even when it’s done by the original artist!
The new algorithm was developed by Milan Rajkovic and Milos Milovanovic, who say that machine vision techniques are now advanced enough to detect the tell-tale signs when a work of art is copied.
Computer identifies copied works of art
The pair of scientists commissioned a Dutch artist to make seven paintings, and then copy them as closely as possible. They then looked at the red, green and blue visible light wavelengths using wavelet analysis – where the computer looks at progressively more blurred images.
From the analysis the computer was then able to tell which paintings were more complex: “for all patches and all the paintings, the mean global complexity of an original painting is larger than the corresponding value of a copy”, they said.
The two then used the technique to examine near-identical paintings that were both created by the 20th century artist René Magritte and have been puzzling art historians for years. They’ve identified one as the original and one as a copy – but they won’t say which!
The pair published the details of the algorithm in a paper which can be found here.