nancy_gohring
Writer

Microsoft photo technologies aim big

news
Jul 18, 20073 mins

Redmond's research arm shows off technologies from innovative photo projects

Microsoft is working on a variety of innovative photo projects, ranging from experiments with its 3-D maps offering to massive panoramic photos that users can zoom into for details.

Developers who work in the company’s research arm showed off the technologies on Tuesday during the Microsoft Research Faculty Summit in Redmond, Wash.

HD View is one photo project that definitely has the “wow” factor. The technology allows users to combine hundreds of photos to create one massive picture that users can zoom in on to see clear details. In one example, a panoramic photo of the city of Seattle includes 800 images, each 8 megapixels in size, stitched together to create a 3.6 billion-pixel image.

On a computer screen, it looks just like a panoramic photo. So what’s the point of combining so many photos? The massive file includes incredible detail.

Michael Cohen, a researcher at Microsoft working on the project, zoomed in to the roof of a building where a clay owl peers around a corner. With the picture zoomed out, a viewer doesn’t even see a pin prick in the spot where the owl sits.

Another large photo of a mountain in Canada looks like a standard nature snapshot. But Cohen zoomed in to discover that climbers are scaling the rock wall. After finding the first climber, he followed the climbing ropes up to find the second one above him on the wall. When the photo is zoomed out, it’s hard to imagine there might be climbers on the wall.

Microsoft offers the tool to build HD View photos for free on its Web site. Creating an HD View panorama image, however, isn’t for just anyone. Such images are quite large and may require special cameras.

Another Microsoft project, unveiled last year and built in collaboration with the University of Washington, collects images of a site such as Rome’s Trevi fountain from public photo-sharing Web pages such as Flickr. The Photo Tourism technology combines the photos into a 3-D image so that users can look at the object from any view. The idea was to take advantage of the potentially billions of images that are online, said Noah Snavely, a researcher at the University of Washington who works on the project with Microsoft researchers.

Microsoft also demonstrated at the summit some experiments with Virtual Earth. Eyal Ofek, a Microsoft researcher, demonstrated a 3-D map of San Francisco that is made up of 10 million images, including 50,000 aerial photographs as well as pictures taken at street level. All the photos are stitched together so that a user can navigate from a bird’s-eye view seamlessly down to street level. The view is different from the street view capability in Google Maps, which doesn’t combine the street-level pictures with aerial shots.

About 800 workers are developing projects at Microsoft Research. Some technologies they develop may become commercial Microsoft products, and others could be sold to other companies. The summit, which ended Tuesday, was an opportunity for Microsoft and its partners in academia to show off some of their projects.

nancy_gohring

Nancy Gohring is a freelance journalist who started writing about mobile phones just in time to cover the transition to digital. She's written about PCs from Hanover, cellular networks from Singapore, wireless standards from Cyprus, cloud computing from Seattle and just about any technology subject you can think of from Las Vegas. Her work has appeared in the New York Times, Computerworld, Wired, the Seattle Times and other well-respected publications.

More from this author