A new software system employing an overhead video camera may soon help you sort and organize the piles of documents on your desk. Find out how this new technology will clear up the paper trail:

Researchers from the University of Washington are bringing the advantages of the digital realm into the physical world–in particular, our workplace. Through a system that utilizes a computer and a video camera, they are helping cure a common condition–the disorderly desk. The technology they’ve developed tracks physical documents and automatically associates them with corresponding electronic documents.

That means you can type, say, ‘HR form,’ and the software will use computer vision techniques to help you zero in on its exact physical location on your desk, even if it’s buried under many stacks of paper. Aside from keywords, you can also use appearance and date of most recent movement to find a document. This is because the system constructs a complete representation of the desk’s spatial structure.

So far, the researchers have built two prototypes–a paper-tracking system and a photo-sorting system. In the photo-sorting application, users physically sort printouts of photographs and the system automatically organizes the digital photographs according to the physical arrangement. This means users will get the best of both worlds, says researcher Jiwon Kim. They get the “tangible interactions” provided by paper and the convenience of “editing, sharing and indexing” provided by the electronic format.

Amazingly, according to Kim, users will not need special tags, paper or marks to employ the new system. Users can “keep using the old paper while only adding a single video camera to the environment,” she says. What’s more, the software can start tackling an already-messy desk. “We didn’t want to force the user to start with an empty desk,” she tells Technology Research News. “Instead, the system gradually discovers the paper documents on the desk over time, as the user moves them around.” Kim worked with research colleagues Steven M. Seitz and Maneesh Agrawala.

The system could hit the market in three to four years. Meanwhile, the researchers are already setting their sights beyond pinpointing the physical location of objects. According to Kim, the system could be refined to enable users to inquire about the history of documents, take written annotations from document surfaces, recognize text and link reminders to documents.

Source:

Video Organizes Paper
Kimberly Patch
Technology Research News, January 12/19, 2005
www.trnmag.com/Stories/2005/011205/Video_organizes_paper_011205.html