Find all the information about our approach at
http://balloon.media.mit.edu/

THANK YOU AND… GOOD LUCK!

The MIT Red Balloon Challenge Team

And we are “cupcake-free”.

Exactly. Why would the Defense Department want to know how quickly a group of Americans could organize themselves to identify and locate a similar event happening in different places around the country? Why would anyone want to help them? All power to the ebay baloon buyers! Unless they’re really trying to see how fast they can dive up the price of an obscure commodity on ebay. Or maybe they want to see how many people will intentionally sabotage a Defense Department project, if only given the chance. Or maybe..

So the one who make noise and know it will be the winner.

If you want to make money, may be you should not try to find the 10 balloon, but sell the location of one to a TV or some media independently of them winning the price, if the story get enough attention they might be interested in being the first media to reveal the location.

But with that money involved i predict a DOS attack on the DARPA website or at least hack attempts. Maybe the DARPA website will resist, but what about the competing websites?
Well maybe that will turn in a Internet security test after all!

Yes how-ever you will possibly have hundreds of images taken from different camera of the same balloon. It would be hard for a computer to know that 100 different images are the same balloon or not. It would only eliminate copies of the same image. Reducing your count by half or so. Still very hard to do. How-ever a keyword search on places like MS, FB, Twitter, etc. might yield better results.

My idea for finding them would be a few carefully crafted google image searches of facebook and twitter for all posts with the words red baloon between when the baloons are put up and taken down, then search the text matches for any locations or landmarks, then using md5 and simple image recognition, remove all duplicate images. Of course, this is well beyond my skill set. I might try anyway using a bash script. (maybe wget and grep will be enough)