Mindboggling Photosynth!

I saw a video yesterday (thanks for the heads up, Tony) that showed a gorgeous application called Photosynth. You can check it out at Microsoft Labs and as the news spreads there will be more videos and articles about it on the internet. If I understand it correctly – image resolution will no longer be a problem.

It creates a three-D image of other images – thousands of images taken by thousands of people. You can zoom in and out without losing quality.

Think back to a time when you have found a picture on the web that you wanted to use in a document. You saved it and then inserted it into your document but it was just a little thumbnail. You stretched it out to the size you wanted and it became ugly, blurry, and pixelated (or blocky) Not so with this – if you watch the demonstration you will see an entire book and the presenter will zoom in and the fonts will stay clear and smooth no matter what size you view.

The three-D part is where is gets confusing and exciting. You have this huge collection of photos (his demonstration used photos of Notre Dame) and all of the view might be slightly different and yet you can use this to get a 3D view of the site. You can also see from different perspectives and it looks like you can see the exact spot the photo was taken from.

Another aspect of this that is very cool is metadata. Metadata is everywhere and it simply means data about data. A non-computer example would be card catalogs in libraries. The card catalog is not the book but it tells you where the book can be found, who wrote it, and other information about the book. Metadata can describe a file or a web page. If you are looking at a web page and view the source code you will see some entries near the top that begin with meta – these usually contain information like keywords that help search engines find the website. Imagine that all these photos have been tagged with keywords and also contain metadata that gives you the time and date the photo was taken, gps coordinates for the exact spot where the photo was taken and much more. Imagine that all these photos coming together as a 3D view of the groups of them and the metadata for one becomes part of all.

I’m still trying to digest it. It gives a whole new meaning to research on the web. I think this will be huge and there will be more and better information as this becomes more known. My only saving grace is that in trying to find out more about it I ran across people much more knowledgeable than I am who were also asking tons of questions.

The implications for ecology, biology, and even astronomy are interesting. Of course we will needs some travelers willing to take their digital cameras to space for that. Imagine using this with microscopic photography, pictures taken from inside hurricanes and tornadoes, legal implications if this type of software is able to create images that can be admitted as evidence in a case.

What about the cultural implications of this and all the metadata that exists out there that leaves our “footprints”. If this application can zoom in on a building detail it can pick out the details of a face or hand in a crowd. We have historically worried about protecting our “privacy” on the internet but I wonder if as we become more aware of how transparent we are and how discoverable, will it cause us to live differently? You know the old saying about people who live in glass houses shouldn’t throw stones? They shouldn’t walk around naked with the lights on either. However – if you know the lights are going to be on – won’t you behave differently? Maybe this transparency isn’t all bad – if it means that we will strive to be perceived as better then we will act better and in the act, maybe become better. As our culture affects how we develop technology we more and more can see how technology affects culture. This is another one of those times when we can’t see ahead to what the long term repercussions will be but this is one that makes me hopeful.

Visually speaking, some thoughts were expressed in this quote “I used to live in Arizona, several years ago, and you’ll notice that the further away you get from a mountain, the more its figure stays the same regardless how far left you go, or how far right. I wonder how the software handles perspective, and distance on this magnitude.
Buildings, which have unnatural forms will look rather different depending on where you are. But mountains tend to retain their figure when you are further away, walking left or right.” http://channel9.msdn.com

Another person on this particular forum brought up the cameras that they use for intricate surgery and how they could use this technology to build 3D surgery so they could study it further.

Blaise Aguera y Arcas is the presenter and worked on the project before Microsoft purchased the companies that developed it. I hope you will watch the video because my explanation and description do not come close to the reality of this application. It truly has to be seen to be believed.

You can see a video here