Evolution Robotics in partnership with Bandai Networks Co. Ltd. recently announced a new ‘Visual Pattern Recognition (ViPR)’ software that allows you to take a picture of an object and email it to a ViPR server to receive information about it, including links to iTunes (for music), YouTube (for video), and the web. The software—which will be offered on the new Spring ‘08 line of KDDI corporation camera phones—identifies distinct and recognizable areas in the select image and compares them to a remote database server, which houses a catalogue of existing pattern specimens.

ViPR is able to learn new objects and images without the need for any special encoding, such as barcodes or watermarks, so that users can take new pictures of objects, images, videos or even locations and tag them with links and content to expand the database. User-generated content will then show up in the results returned to other users who take similar pictures, creating an extensive world-wide visual database for communities to develop and access.

Fortunately, ViPR reportedly performs well on low cost components such as the cameras used on most mobile phones today, even when lighting and other visual conditions are poor. According to iPhone Buzz, Evolution Robotics claims an 80-100% success rate for visual searches.

Here is a pretty nifty demonstration of Evolution Robotics’ visual search technology running on Apple’s iPhone:

Macworld reports that Evolution Robotics is developing a native iPhone application, which they say is due out in June. Because Apple’s iTunes media store is already equipped for wireless purchases and downloads and has a large number of preview tracks and artist information, ViPR and the iPHone would seem an ideal match.

It is not hard to imagine how this software might be used in an academic setting for library searches, building/landmark identification, textbook purchasing, etc.