An interesting news story was reported this week in the tech press and online. It got some coverage, but certainly no bells and whistles and few of those reporting it examined what the potential of the story actually meant.
Perhaps the format of the story was a bit boring. “Google has been awarded a Patent for automatic recognition and tagging of objects in video”. A bit geeky perhaps. Maybe many even thought it already existed? Others maybe not really understanding what it meant? Some not bothered by it all. Tagging objects in videos – not exactly loaded with ‘wow’, but in five years time we think consumers across the globe will be realising just what a wow moment it is.
What does it mean?
Well think Facebook and how you tag your mates in your photographs that you post up. It helps identify who is who, encourages sharing and is at the heart of the ‘social’ aspect of Facebook. Under this patent Google would own the rights in the technology to automatically tag things, in moving video. By things we mean lamposts, cars, but also people. There may also be the possibility to identify specific people using facial recognition technology, although that’s not immediately part of this patent.
Beyond the Patent
The key probably isn’t in the Patent itself – although with Google the owner of the biggest video database on the planet in the form of YouTube the ownership is a massive coup for them – The key is how it could potentially link up with other technology. Not even technology yet to be developed, but stuff that is around already, maybe cutting edge but still affordable and coming into more every day use. So facial recognition technology. Already a number of sub £100 point and click digital cameras have the basics built in and linking that technology with a database of face shots of your friends and family could enable your camera to identify people and tag them direct in still images for you. The same would be true with video under this latest Google Patent. So what if that image recognition was then linked with Augmented Reality? The kind of thing that Aurasma and Blippar are putting to great use with various brands and advertising campaigns? With augemented reality you can already overlay different content over an image that your smartphone recognises. So what if that could be combined with video and object recognition? Think Photoshopped photographs and stick that in a moving image format… Effectively, automatic recognition could identify the lead character in a film and overlay a different person into that moving image. Hey presto…..Batman: The Final Return or Star Wars: Another Generation.. the blockbuster movies of the future could soon be fully interactive on the home stage, enabling you to feature in your own favourite films.
Of course, the agreement of the big studios to that – even if possible – would likely take decades to complete, but certainly within the next decade we would expect to see technology of that kind in use for your own home movies, in semi-professional stuff streamed online and more gimmicky professional stuff too. If you can already use augmented reality to star in your favourite football team’s pre-season photograph, then being able to star in their round-up of the season video is probably not too far away either.