As mobile phones become more sophisticated, they are bidding to replace laptops as the business traveler's tool of choice. But trying to view and navigate documents on a phone's small screen remains frustrating. A new research project at Fuji Xerox Palo Alto Laboratory (FXPAL) aims to solve that problem, while making it easier to transfer scanned documents to a phone in the first place.

In a recent demonstration, researchers showed how the technology, called Seamless Documents, could store a scanned document in a database and analyze its structure and content. The analysis identifies sections and paragraphs, and automatically extracts key phrases that summarize the sections. So when a person pulls up the document on a phone, she can jump to a section labeled with a keyword, or just skip to the last paragraph on a page. In addition, as the user scrolls through the document, software on the phone automatically resizes images, section headers, and plain text, as different elements of the layout come into view.

Thanks to Moore's Law, mobile phones have been gaining exponentially in computational power, but their screen sizes have remained comparatively static. Researchers, such as Patrick Baudisch at Microsoft, have been trying to find the best way to modify the mobile user interface for viewing Web pages and maps. And Apple has made great strides in making content viewable on its iPhone. But no one has solved the complex problem of building software that lets a user skim through and zoom in and out of scanned documents without losing her orientation.

Solving this problem could be a boon to business travelers who now lug around heavy folders of papers. "Executives take enormous stacks of documents with them when they travel," says Scott Carter, a researcher on the FXPAL project. "We wanted to find a better way for them to get through these documents quickly."

While the FXPAL project is only six months old and still very much a research project, it aims to solve a mixture of problems simultaneously--from extracting information from analog documents and synchronizing it across devices, to making this information easy to navigate on a cell-phone screen.

The first part of the Seamless Documents project focuses on converting analog documents into digital information and storing it in a database accessible over both the Internet and the cell-phone network. Once a scanner or some type of camera captures the documents, they are sent to a database where specialized software, developed at FXPAL, analyzes their structure to determine where paragraph breaks, pictures, and section titles occur. In addition, the document is run through optical character recognition software that converts the images of words on paper into digital information that can be read by a computer. Then, software automatically summarizes the text, picking out key words and concepts from each section that are then highlighted for the user.