Why Crowdsourcing Might Be Better than Secret Service

Matt.Bachman, College Mogul

Microsoft put its Photosynth technology to work yesterday at the Presidential Inauguration and the results are impressive. In a nutshell, the application builds a navigable, 3-D model of a landscape by compiling, sorting, and repositioning a collection of photos. Here’s the kicker: the app is crowdsource-capable. (What is crowdsourcing? ) Working with CNN photographers located among the masses, Microsoft successfully mapped the scene around the Capitol building. By allocating each photo to a specific point cloud, or notable point of focus within the photo, the application calculates patterns and variations in the photo cloud, locates the vantage point of each photo, and finally positions it accordingly within the model.

See Photosynthesis

The decision to showcase technology calculating vantage points and mapping a seemingly infinite area at an event where the President is exposed to a massive crowd underscores the security potential of Photosynth. It is not hard to imagine Presidential security using this software to track a sniper hidden within a sea of people. More astounding is that the application’s performance improves with increasing scale, as more photos create a more detailed model.

Photosynth is just one of several surveillance technologies that rely on a human network to provide near-real time information and become increasingly effective as they scale. Back in October, CM profiled Purdue University’s Distributed Nuclear Detection by Ubiquitous Cell Phone project, where a chip embedded in a network mobile devices can detect radiation from a dirty bomb and triangulate its location. More recently, the surveillance capability of mobile phones allowed users to almost instantly upload photos of flight 1549 from passing ferries and midtown offices to Twitter —before mainstream media or authorities reached the scene its crash in the Hudson.

More