 |
Efforts to put information on the Web face many obstacles
Monday, October 25, 1999
By RUTH SCHUBERT
The line of red light cutting across David's nose was an effort to preserve, not destroy, Michelangelo's towering ideal of male beauty.
The goal is to create three-dimensional geometric models of 10 Michelangelo masterworks. The computer models, detailed down to the chisel mark, will be available over the Internet to researchers around the world.
The ambitious goal of digitizing some of Florence's most treasured sculptures offers a tantalizing glimpse at the vast repository that may some day be available on the World Wide Web -- and a cautionary view of the limits that must be overcome.
More books and artworks, music and manuscripts and films and photographs are being digitized and placed on the Web than ever before.
But the process is hindered by copyright issues. Better technology is needed to scan, store and transmit the huge volumes of digital information. And users need better ways to search multimedia repositories for the information they need.
Many see a huge incentive to overcoming those obstacles to create a digital library unlike anything on Earth.
"This is one investment that benefits every man, woman and child in the world," said Raj Reddy, a computer science and robotics professor at Carnegie Mellon University whose research encompasses universal digital libraries.
"That's what is the exciting aspect of it. It's not just for the benefit of a select few, but it's for the benefit of everyone. What that means is people like you or me who don't have a Harvard card can have access to a library bigger than the Harvard library."
Six federal agencies, including the National Science Foundation and the Library of Congress, are sponsoring a $40 million to $50 million research initiative aimed at taking digital libraries to the next level. They'd like to make more information available and create technology that makes the information more flexible, dynamic and easy to use. Called the Digital Libraries Initiative, the project is also sponsored by the Defense Advanced Research Projects Agency, National Library of Medicine, National Aeronautics & Space Administration and the National Endowment for the Humanities.
Many other efforts also are under way at colleges, universities and institutions around the country, including the University of Washington in Seattle.
The Michelangelo project, funded by the Digital Libraries Initiative, was designed primarily to test technology. The David was selected largely to create a stir.
The work, he says, allows art historians to "see the mind of the sculptor."
It took two years for a team of 30 faculty, staff and students to design, build and test the custom laser scanner and mechanical gantry used to record the surface of the David, which stands 23 feet tall on its pedestal.
Recording the billions of bits of information was no simple process. A laser was aimed at the surface of the sculpture. Then, a camera placed at a right angle to the laser would pick up the surface line reflected by the laser's light. By merging together thousands of these single recorded surface lines, the computer is able to form a fine mesh that recreates the surface.
The digital David the team created is formed from 2 billion polygons and 7,000 images.
The project uses Stanford-developed programming that's pushing the boundaries of technology. But the sheer volume of data collected prevents the researchers from creating the highest theoretically possible resolution.
"That's starting to break our tools," Curless said. "We have another sort of order of magnitude under the hood that we haven't been able to reach."
The two dozen grants awarded as part of the national Digital Libraries Initiative include several projects that are attempting to move beyond straightforward scanning.
Three researchers at the University of Kentucky, for example, are trying to use technology to restore severely damaged manuscripts in the Cottonian Collection at the British Library. The historical documents amassed by Sir Robert Bruce Cotton, 1571-1631, are considered the most important collection of manuscripts ever assembled in Britain by a private individual. Surpassing that of government at the time, Cotton's library laid the foundation for public records laws and included the first public law library.
Using special lighting and two- and three-dimensional scanning, the researchers hope to recover information invisible to the naked eye to create a digital restoration.
Faculty at the University of Texas in Austin, in collaboration with research universities and natural history museums around the world, are using high-resolution X-ray computer technology to create a library of 3-D vertebrate skeletons. The study encompasses skeletons in all forms, from fossils to embryos to living adult animals.
John Kappelman, one of the principal investigators at UT, notes that recent advances in three-dimensional digitizing hardware and software go "far beyond digitizing traditional materials such as hard copy text, images and video." The result, he says, will "significantly alter the definition of a national library."
At UW, a dozen "special collections" are currently online, including the Cities and Buildings Collection, an archive of more than 4,000 buildings around the world.
Assistant Professor Alex Anderson, who is contributing to the architectural collection, also links to it on his course Web site.
In class, he uses a traditional slide show to illustrate the architecture of the ancient world. By next year, he hopes to have the digitized collection available for classroom use.
Student Lucas Clara, a UW junior majoring in international studies, appreciates the Web page link, but says the collection needs to be even more comprehensive and have a better search function.
Most of the other special collections available at the UW cover the history and culture of the Northwest. The library hopes to bring them together in a Pacific Northwest library of maps, guides, photos, diaries, manuscripts, moving images, oral histories and instructional slides that will be available for school classes and the public at large.
"It's really exciting for libraries because we're beginning to see a resurgence of interest in primary documents," said Geri Bunker Ingram, coordinator of the UW Libraries' Digital Initiatives Program.
The biggest stumbling block, she said, is working through the copyright and permission issues that surround digitizing and disseminating materials placed in the UW's care decades ago.
"It's not the technical stuff that's the problem, it's the legal stuff," Bunker said. "We're making it up as we go along."
Currently, the UW Libraries are creating the "American Indians of the Pacific Northwest" collection, in partnership with Seattle's Museum of History and Industry and the Eastern Washington State Historical Society. Composed of 2,500 photos and 6,000 pages of text, the resulting digital collection will be added to the Library of Congress' American Memory Project.
The American Memory Project is the Library of Congress' effort to begin digitizing resources that chronicle America's history and heritage. At present, 38 collections are on the Web, ranging from documents, photos, films and recordings from America's vaudeville era, to films of San Francisco before and after the earthquake of 1906, to American sheet music from the 19th century.
There are two schools of thought over where the focus should be. Some experts favor the rarest, special collections; others are focused on getting as much as possible up on the Web as quickly as possible.
"Our attitude is that eventually everything will be digitized so it's not a matter of what, it's a matter of what first," said Mike Shamos, who is working on the Universal Library Project at Carnegie Mellon University.
The Universal Library is one of the broadest attempts at putting material online. But it hasn't been simple. Shamos tells of approaching a museum that holds one of the world's largest collections of illuminated manuscripts in the world. The curator turned down his offer to digitize the material.
"The people that have the good stuff are extremely jealous about having them shared," Shamos said. "Most of them turn us down flat."
Another Carnegie Mellon project, called the Informedia Digital Video Library, mixes technology that recognizes speech, languages and images to automatically transcribe, segment and index news and documentary broadcasts. They currently have more than 1,500 hours of material.
The researchers see the technology benefiting not just libraries, but also news organizations, political organizations and many others.
As a counterbalance to those who see publishers jealously guarding copyrights and libraries keeping the Gutenberg Bible under lock and key, some see simple economics pushing more information onto the Web.
"I think in 20 years most things will be available on the Web, and it will be available much cheaper than it is now," said Jim Gray, a senior researcher at Microsoft Research in San Francisco.
"There's huge economic savings in delivering on demand. It's the ultimate in reversing the charges. The person who's going to be downloading the book is going to be paying all the printing costs, all the distribution costs. They're even paying for the phone call."
Dan Weld founded NetBot. Oren Etzioni created Metacrawler. So it's not surprising that when the two University of Washington computer scientists sat down to write a research proposal for the National Digital Libraries Initiative, they were thinking search engines.
What they come up with is a project to create a kind of virtual reference librarian, a Web tool that would be smart enough to go out and query only the most appropriate search engines.
"If you have a query about a particular stellar body, a quasar, say . . . it doesn't make any sense to send your query to the Shakespeare or the nutrition server," Weld said. "Given a query, we would like our engine to figure out precisely which of these sources to send the query out to. . . . We're trying to build what we call 'reference intelligence.'"
P-I reporter Ruth Schubert can be reached at 206-448-8130 or ruthschubert@seattle-pi.com
SEATTLE POST-INTELLIGENCER REPORTER
Working in the after-hours quiet of Florence's Galleria dell'Accademia, a team of computer science researchers from Stanford University spent a month earlier this year laser scanning every quarter-millimeter of the David.
To create a digital "David," a laser was aimed at the image, and a camera picked up reflected surface lines. In all, 7,000 images and 2 billion polygons created the image.
"We wanted to have high impact, we wanted to be highly visible," said Brian Curless, an assistant professor at the University of Washington who was a graduate student at Stanford when the project began.
Brian Curless, today an assistant professor at the University of Washington, was part of a Stanford project that wanted to show the possibilities of material that can be placed on the World Wide Web by figuring out how to post a three-dimensional image of Michelangelo's "David" on the Internet.
Paul Joseph Brown/P-I
A virtual librarian
Related web sites
>>
The Digital Michelangelo Project
>>
University of Washington Libraries
>>
Library of Congress American Memories Project
>>
The Universal Library at Carnegie Mellon
>>
The Jack London Collection
>>
National Gallery of the Spoken Word
· Web of support buoys presidential candidates
· Public Health suing two bars for letting customers light up
· 'Nickelsville' moves to U District
· Calif. man charged in sex crime spending time in Seattle
· Measuring McCain as a political 'maverick'
· Feds link prescription drug ring to Poulsbo
· Burien man who killed 3 for money sentenced to 97 years in prison
· Sham marriage ring for stocking brothels alleged
· Mother charged with abandonment
· Neighborhood funds program marks 20 years
· Settlement reached in fatal crane collapse in Bellevue
· City considers buying Vancouver newspaper building
more
· Today's News: Our Take - Angelina Jolie Breastfeeds Twin on W Magazine Cover
· Sham marriage ring for stocking brothels alleged
· Alaska panel finds Palin abused power in firing
· NY county sends ballots with Barack 'Osama'
more
101 Elliott Ave. W.
Seattle, WA 98119
(206) 448-8000
Home Delivery: (206) 464-2121 or (800) 542-0820
seattlepi.com serves about 1.7 million unique visitors
and 30 million page views each month.
Send comments to newmedia@seattlepi.com
Send investigative tips to iteam@seattlepi.com
©1996-2008 Seattle Post-Intelligencer
Terms of Use/Privacy Policy
