Microsoft’s Photosynth a huge step in visual technology
Samuel Van Eerden
August 15, 2006
Technology is growing in huge strides. The “geek” is no longer an unloved member of society. He may be mocked and teased for his bug-eyed glasses and unkempt hair, but underneath the stereotype, he is revolutionizing the world, one kilobyte at a time.
And the rest of us marvel.
Sometimes, though, the continuous flow of innovative new products and applications is so torrential, it becomes almost mundane. Even the amazing is normal if the product is joined by so many other amazing creations. In this age of (often open-source) technology, it is sometimes difficult to make any one invention stand above the competition.
But suddenly, a program named “Photosynth” has appeared on the scene and done just that: elevated itself above the rest by being not only a brilliant design, but by forming a benchmark that others can work off of to go above and beyond, using Photosynth’s technology.
I never thought I’d say this, but Microsoft (in conjunction with the University of Washington) is at the head of the class with this one!
According to Photosynth’s official webpage (http://labs.live.com/photosynth/), the program is a “new technology from Microsoft Live Labs [the team’s first product release] that will change the way you think about digital photos forever. Photosynth takes a large collection of photos of a place or object, analyzes them for similarities, and displays them in a reconstructed 3-Dimensional space.”
This is basically Google Earth to the next level. Where Google took satellite images and allows users to maneuver seamlessly over the world’s landscape in various levels of image resolution, Photosynth takes this idea to the extreme, and brings the camera to ground-level. Google Earth’s capabilities are good, but they are still limited to how close they can get you to the earth’s surface.
The idea of Photosynth is to take any and all available digital photographs of any particular location and compile them into a single 3-D image. The program (available for download later this year) will make an exhaustive (continuously updated) database of online and user-submitted photos which can be easily accessed by anyone with the program.
Part of the brilliance of this technology lies in its ease of operation. Almost everything is automated; the user just tells the computer what location he wants to view, and then goes from there. Photosynth begins by analyzing each photo for identifiable points, and then matches those images with others that have identical points. Say for instance you were at a Green Bay Packer football game at Lambeau Field and you took a bunch of pictures of the stadium. When you upload them to Photosynth, the program will instantly see where they all line up and then arrange them in their actual order. In this example, let’s say you took ten pictures of the playing field. Obviously there is some overlap there (ie: the 50-yard line shows up in five of those photos), but Photosynth takes all the pictures and puts them together, like puzzle pieces until you have one continuous image you can virtually “walk” through. There is also the option of zooming out and seeing all the individual photos, with the choice of selecting one in particular. Left-click and drag over the picture, and the cursor creates a box which will zoom in however near you wish (as long as such a close-up photo exists in the database).
By this time I’m sure you can see the astounding implications of this program, but let me lay it out for you, further.
We’ll say you just returned from a trip to Rome where you toured the Coliseum and the Catacombs. You’ve taken a lot of pictures, which you upload to Photosynth. Photosynth adds your photos, then searches its database for images that have already been taken of the same locations. After all, Photosynth is designed to be a giant compendium of the photography collections of the entire world! Literally. (Or at least whoever owns the program.) That is an amazing concept! Now you can VIRTUALLY tour the Coliseum and the Catacombs at any time, from an eye-level perspective, with the ability to zoom in or out almost to infinity.
The possibilities are staggering; the implications wild and unprecedented. Digital cameras are becoming increasingly popular, and with them, online photo albums like Flickr, Photobucket, and Snapfish. Once Photosynth is released, you can bet that its popularity will grow exponentially.
And then it’s only a matter of time before you can walk the streets of Paris, tour the Vatican, and visit Yellowstone National Park . . . all from your own home.
Take it a step further and forget about national landmarks like the Eiffel Tower, the Taj Mahal, or the Great Wall of China. What about the Rainforest or your own bustling metropolis (be it New York City or Gary, Indiana)? Next on the agenda for Photosynth is to get 3-D mapping of the INSIDE of buildings. Imagine being able to check out a restaurant or shopping mall before actually going there.
Obviously, in this world of terror and violence, this kind of technology can be seen as a huge threat to our national safety. I see that discussion as another article altogether, and since mine is more to inform then to commentate on the privacy concerns involved, I’ll leave that debate for another time.
Suffice it to say, Microsoft has discovered something truly remarkable, and I—for one—eagerly await Photosynth’s release.
(Read the official paper on Photosynth written by Noah Snavely, Steven M. Seitz, Richard Szeliski: http://phototour.cs.washington.edu/Photo_Tourism.pdf)
And the rest of us marvel.
Sometimes, though, the continuous flow of innovative new products and applications is so torrential, it becomes almost mundane. Even the amazing is normal if the product is joined by so many other amazing creations. In this age of (often open-source) technology, it is sometimes difficult to make any one invention stand above the competition.
But suddenly, a program named “Photosynth” has appeared on the scene and done just that: elevated itself above the rest by being not only a brilliant design, but by forming a benchmark that others can work off of to go above and beyond, using Photosynth’s technology.
I never thought I’d say this, but Microsoft (in conjunction with the University of Washington) is at the head of the class with this one!
According to Photosynth’s official webpage (http://labs.live.com/photosynth/), the program is a “new technology from Microsoft Live Labs [the team’s first product release] that will change the way you think about digital photos forever. Photosynth takes a large collection of photos of a place or object, analyzes them for similarities, and displays them in a reconstructed 3-Dimensional space.”
This is basically Google Earth to the next level. Where Google took satellite images and allows users to maneuver seamlessly over the world’s landscape in various levels of image resolution, Photosynth takes this idea to the extreme, and brings the camera to ground-level. Google Earth’s capabilities are good, but they are still limited to how close they can get you to the earth’s surface.
The idea of Photosynth is to take any and all available digital photographs of any particular location and compile them into a single 3-D image. The program (available for download later this year) will make an exhaustive (continuously updated) database of online and user-submitted photos which can be easily accessed by anyone with the program.
Part of the brilliance of this technology lies in its ease of operation. Almost everything is automated; the user just tells the computer what location he wants to view, and then goes from there. Photosynth begins by analyzing each photo for identifiable points, and then matches those images with others that have identical points. Say for instance you were at a Green Bay Packer football game at Lambeau Field and you took a bunch of pictures of the stadium. When you upload them to Photosynth, the program will instantly see where they all line up and then arrange them in their actual order. In this example, let’s say you took ten pictures of the playing field. Obviously there is some overlap there (ie: the 50-yard line shows up in five of those photos), but Photosynth takes all the pictures and puts them together, like puzzle pieces until you have one continuous image you can virtually “walk” through. There is also the option of zooming out and seeing all the individual photos, with the choice of selecting one in particular. Left-click and drag over the picture, and the cursor creates a box which will zoom in however near you wish (as long as such a close-up photo exists in the database).
By this time I’m sure you can see the astounding implications of this program, but let me lay it out for you, further.
We’ll say you just returned from a trip to Rome where you toured the Coliseum and the Catacombs. You’ve taken a lot of pictures, which you upload to Photosynth. Photosynth adds your photos, then searches its database for images that have already been taken of the same locations. After all, Photosynth is designed to be a giant compendium of the photography collections of the entire world! Literally. (Or at least whoever owns the program.) That is an amazing concept! Now you can VIRTUALLY tour the Coliseum and the Catacombs at any time, from an eye-level perspective, with the ability to zoom in or out almost to infinity.
The possibilities are staggering; the implications wild and unprecedented. Digital cameras are becoming increasingly popular, and with them, online photo albums like Flickr, Photobucket, and Snapfish. Once Photosynth is released, you can bet that its popularity will grow exponentially.
And then it’s only a matter of time before you can walk the streets of Paris, tour the Vatican, and visit Yellowstone National Park . . . all from your own home.
Take it a step further and forget about national landmarks like the Eiffel Tower, the Taj Mahal, or the Great Wall of China. What about the Rainforest or your own bustling metropolis (be it New York City or Gary, Indiana)? Next on the agenda for Photosynth is to get 3-D mapping of the INSIDE of buildings. Imagine being able to check out a restaurant or shopping mall before actually going there.
Obviously, in this world of terror and violence, this kind of technology can be seen as a huge threat to our national safety. I see that discussion as another article altogether, and since mine is more to inform then to commentate on the privacy concerns involved, I’ll leave that debate for another time.
Suffice it to say, Microsoft has discovered something truly remarkable, and I—for one—eagerly await Photosynth’s release.
(Read the official paper on Photosynth written by Noah Snavely, Steven M. Seitz, Richard Szeliski: http://phototour.cs.washington.edu/Photo_Tourism.pdf)