Who needs QR Codes when all it takes is a picture?

2 minute read

As I anticipated in my post “Bridging the physical barrier: QR Codes, NFC and AR”, one of the upcoming ways to interface the physical world to the digital world is coming from the field of augmented reality, of in other words all the researches that are trying to create algorithm able to “detect” from a camera what’s happening in the world and interact with it.

Renzo Giust and Clément Delangue (Moodstocks)  pointed me to a few resources, some known or quite famous, and some others less so:

  1. Layar Vision: the company that was building mobile apps to add layers to the camera input, now has released the Vision feature, where you can allow your customers to point the camera to any object you want and it will detect it. Form Android and iPhone.
  2. Moodstocks Video SDK: an API to allow developers to add real-time object recognition in the apps “in minutes” with also an analytics platform. For Android, iPhone and Blackberry.
  3. Kooaba Paperboy and DejaVu: two apps that detect images you shoot and links it to its digital equivalent. Detects books, cds, dvds, wines, etc. They also sell their tech API to do the same, Smart Visuals. For Android and iPhone.
  4. LTU engine: an engine to detect images given a specific fingerprint. I wasn’t able to get more details about it from the website, but they seem selling the technology layer.
  5. (do you know others?)
Of course this is just the beginning and it’s probably going to require either the same feature at OS level, provided directly by iOS or Android, or a really simple opensource library to drive the adoption of this technology, however, we are already at the point where:
  • It’s delivered at low cost to consumers
  • It works in real-time on a smartphone

I don’t know yet what are exactly the limits of these technologies, as often happens the devil is in the details. From what I’ve seen until now probably they are currently limited to work on a specific subset of images, and that means that while they are simpler than QR Codes by not requiring any specific element on the page, they probably can’t work as “generic” readers, since a matching must exist somewhere else. Unless, of course, it’s Google or someone like that providing the search service or they get smart enough to read text and URLs automatically. ;)

I think that these technologies are already a good alternative to QR Codes in some specific scenarios and they might be even more engaging because they feel even more like magic to the final user.