Posted by: crudbasher | May 15, 2012

What is Computational Photography?

In 2007 Google launched a new feature of Google Maps called Street View. With this, users can see 360 degree panoramas from various points on a map. These images were generated by a fleet of cars equipped with camera and GPS technology. At first the images were only of a few cities but it has spread to much of the US and even some foreign countries. I imagine this must have been a substantial investment of time and money for Google but it is a very cool idea.

One of my influences is Ray Kurzweil. He’s a futurist and inventor who came up with something called the Law of Accelerating Returns. It basically is an observation that the rate of technological change has been accelerating over time. For example, he tells this story about the genome project.

In a 2009 TED talk on his theory of singularity, Ray Kurzweil used the example of the Human Genome Project (sequencing DNA) to explain the path and power of exponential growth. Started in 1990, HGP began the process of mapping the sequence of the genome, which proved more difficult and costly than first anticipated. After 1/2 the projected time to complete the project was over (completed in 2003), critics of the project complained that HGP was only 1% complete with the sequencing. However, due to power of exponential growth, HGP was able to finish on time because the remaining work did not take a linear path to completion. Researchers were able to double the 1% seven more times in the remaining 6.5 years and finish on time. (H/T Engaging Change)

I came across this story today about a new project using the billions of photos stored on the Internet. The research involves using photos freely shared online and combining them to produce information. For example, this image from their website shows how they use multiple photos of the Coliseum in Rome to construct a 3d model of the structure.

While I have seen this sort of technology before, these folks have taken it to the next level. They are using the embedded timestamp in most photos to be reconstruct a person’s movements. If you know where each photo was taken and when, you can see where they moved, and how fast. If you have a sufficiently large number of photos you can then determine traffic patterns over a large scale. This sort of research was never practical until now.

Human Movement Patterns In Manhattan

As I mentioned previously, this is The Year Of Big Data. So how does this apply to Google Maps  Street View? Like the Genome project the technology is advancing very quickly so it’s likely the next generation of Google Street View photos will be taken by us. This data will continue to be exploited in ways we don’t even understand yet. This data analysis will even come to the school system.

The Internet is leaving the purely virtual world and is entering the physical. We live in amazing times!


  • Modeling People and Places with Internet Photo Collections – ACM Queue
    • Photo-sharing sites such as Flickr and Facebook contain vast amounts of latent information about our world and human behavior. Our recent work has involved building automatic algorithms that analyze large collections of imagery in order to understand and model people and places at a global scale. Geotagged photographs can be used to identify the most photographed places on Earth, as well as to infer the names and visual representations of these places. At a local scale, we can build detailed three-dimensional models of a scene by combining information from thousands of two-dimensional photographs taken by different people and from different vantage points.
    • In aggregate, and in combination with nonvisual metadata available on photo-sharing sites (including photo timestamps, geotags, captions, user profiles, and social contacts), these billions of photos present a rich source of information about the state of the world and the behavior of its people.

Posted from Diigo. The rest of my favorite links are here.




  1. […] A limitation of this technology is it is short range. If you can do it straight from image analysis though it would work at any distance. With several people from different angles you can get a very accurate model. (see What is Computational Photography) […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: