It is just a question of time. Not years but probably months. Already a lot of desktop consumer based application are offering a glimpse of its power : Image recognition. Google has recently taken it a step further. Search with Image recognition.
“The concept of Google Goggles is dead simple- a user snaps a photo of an object around them, be it a book, building, text or any other object, and the app will return a search result tailored to for that object” explains mobilemarketingwatch.com.
Say you walk in front of a building and would like more info, just snap a picture of it, submit to Google Goggles and voila. Same goes for CD’s , Books, paintings and maybe photographs too. Still in its infancy, Google Goggles , when launched live, will mark the debut of image recognition as a replacement of text search. For now, the Goggles project is limited to objects. Soon, once the Google lawyers can figure how to get around privacy laws, it will also work with people too. Meet someone new. Take a snap and in 5 seconds, you will know there whole history. Well, at least the one available online.
What does it mean to the photo industry ? Many, many changes. First, a new way to search. All database will soon be able to also offer such a search to their clients. Google tends to create standards. It is nit a bad thing, except for the technophiles out there. Just imagine: Your metadata could be incorrect and image would still be found. A few companies already offer this technology but not as deep as Google. LTU tech or Idee, for example, will do image matching . That is, if images look similar, regardless of what is in the image. Goggles recognizes what is in the image and looks for that, and not similar images.
Google new visual search will mainly be a huge help for image key wording. Since its inception, it has been notorious for offering free API’s ( that’s a little backdoor that lets two applications talk together) to its technologies. That will allow savvy programmers to tie an image database to Goggles and have all its image indexed automatically. Those microstock will be all over this technology. While certainly incapable to add emotion or concepts, it will however facilitate a lot of the tedious work of recognizing content.
Companies like Imense who is currently offering similar technology will certainly take a huge hit. Same goes with Idee and its CD or book Iphone apps.
While there is nothing dramatically new in this technology, it is the shear power of Google already indexed document that will make this a success. They will be hardly any images that will not return a result, besides maybe those “artistic” ones.
If you have an android phone, you can try it now : Google Goggles.