Science Masters Art

Dennis Faas's picture

Computers can now recognise the work of different artists and even tell the differences between genuine paintings and fakes.

Scientists at Israel's University of Haifa have developed a mathematical program that analyses paintings geometrically. The computer breaks down the picture into its component sections, then calculates the sines and cosines. (If you've forgotten your high school trigonometry, that means the relative size and shapes of triangles.)

Once the computer learns the pattern of a particular artist, it can identify whether another painting is legitimately that person's work. Professor Daniel Keren, who developed the program, explained: "As soon as the computer learns to recognize the clock drawings of Dali, it will recognize his other paintings, even without clocks. As soon as the computer learns to recognize the swirls of Van Gogh, it will recognize them in pictures it has never seen before."

According to Professor Keren, human vision is still a better judge because it "has undergone evolution of millions of years and our field is only 30 years old". He explains that computers are much better at recognising and processing highly intricate three-dimensional images such as the brain's arteries or a road network. But humans have the advantage with seemingly simple tasks such as recognising a human face (Source: University of Haifa)

It seems this is because computers can only deal with specific detail and lack the human ability to see the big picture. Professor Keren says the computer only recognises artists by the common patterns to how they paint, rather than what the image actually is. "It can't yet look at a painting as a whole, and say, this is characteristic of a certain artist because it represents a particular subject -- a cow, or a flower, for example. (Source: vnutnet.com)

Art dealers have already made enquiries about using the program for investigating potential forgeries. But it seems unlikely the computer will replace human art experts in assessing whether an unknown painting is by a particular artist. Instead it could be best used for judging how accurate a billed 'replica' of a famous work is. (Source: vnunet.com)

The project is a great example of the benefits and limitations of technology. The computer is far more capable of objectively analysing a painting, but only a human can make a subjective judgement about its appeal.

Rate this article: 
No votes yet