by Andrea Orr
PALO ALTO, Calif., Jan 29 (Reuters) - Enter the name "Michelangelo" on any Internet search engine and you will be directed to images of David, the Sistine Chapel and other masterpieces that have been posted on online museum sites.
What you will not find is anything like an exact reproduction of the Italian master's work. Virtual art, like cybersex, is still a crude imitation of the real thing.
But computer scientists at Stanford University say they have found a way to replicate famous sculptures with a new degree of precision, capturing a statue from all angles with all its surface details down to the most minute chisel marks.
Stanford Computer Science professor Mark Levoy and a team working on the Digital Michelangelo Project are spending the winter in Italy using a specially designed laser scanner to produce computer images of all of Michelangelo's statues.
Levoy hopes the images they create will contain as much detail of David and other sculptures as an MRI machine captures of the human body. One reason he is using a laser scanner is that no MRI machine is big enough for Michelangelo's statues.
The Digital Michelangelo Project's goal is to build computer models accurate enough to be of use to art scholars as well as casual observers. Levoy contends that none of the computerized statues created to date have had enough detail to be used in serious research.
"A lot of art historians are interested in this data," Levoy said in a telephone interview from Florence. With a digital image, he said, students would be free to turn the work literally on its head or examine it under different lighting conditions, gaining a new perspective on the sculpture.
Eventually Levoy's team hopes to use digitized information to "virtually restore" damaged art works such as the beard on Michelangelo's Moses. They also might animate some sculptures to help art teachers illustrate tricks of perspective.
In the Pieta, for example, Michelangelo solved the problem of fitting a grown man neatly into a woman's lap by giving Mary extremely long legs. Using animation to have Mary stand up could show this distortion more plainly.
The best way to visualize the digital imaging process is as a very high-tech three-dimensional office copier. The laser scanner sends a sheet of light across the sculpture while a video camera looks at the statue's profile and records the shape of the light as it sweeps along the surface.
By scanning the sculpture from several different angles and analysing the changing shape of the light, scientists can create a high-quality reproduction.
Levoy is also applying this technology to an even more daunting project, attempting to put back together the pieces of the Forma Urbis Romae, a 70-by-50-foot (22-by-15 metre) marble map of the Roman empire that fell off a wall and crumbled into thousands of pieces. It has stumped historians for centuries.
The map, long considered a missing link in understanding ancient Rome, has been especially hard to reassemble because less than half its pieces survive and many of them weigh as much as 100 pounds (45 kg) and cannot be easily shuffled around.
But Levoy believes the laser scanning process could present a solution. Most of the remaining fragments are thick enough for him to make digital images of the broken surfaces in order to gather three-dimensional clues for fitting them together.
His efforts with the Forma Urbis Romae earned him a place last year in Wired Magazine's "Wired 25" list of individuals deemed to be "attempting the impossible."
Perhaps more possible is building an extensive database of digitized sculptures. Levoy, who says Michelangelo has been one of his favourite artists since his days as an undergraduate architecture student, hopes eventually to replicate other famous works as top-quality computer images.
"We are doing it with a lot of detail," he said.
One problem with the project, in fact, could be the degree of detail. Levoy says some of his images will contain several gigabytes, more data than the average personal computer is capable of storing.
For this reason, he confessed, it is "difficult to foresee" exactly how the digitized sculptures will be put to practical use in the near future. "They are so far beyond the state of current computers. I am not sure who in the next 10 or 20 years could even download this thing."