I have recently seen several news items talking of google’s new image search tool which they have aptly named goggles. The basic concept is that you can take a photo of an object and through the ‘goggles’ application it gets sent to the google servers which process the image and identify what it finds in the image; then returning relevant results.
Well known landmarks, artwork or products on a store shelves are apparently now no problem for the software. Although the application does also look for barcodes in order to help identify objects does this not surpass barcode technology. Results currently take 1-2 seconds to return, which in many situations is too long, but in the future could barcodes be replaced by machines which simply identify what they see rather than relying on a particular code.
This day may well be decades off, recognising the Eiffel Tower or a can of Coke is one thing but to separate documents or see whether the shopper is buying Organic or Non-Organic broccoli is likely a step to far. However with the current trend for QR codes in advertising it is not infeasible to suggest that eventually a picture of an advert, with no barcode, could return a webpage when snapped with a mobile since it would only take google linking an image into their database.
So is a project like Google Goggles (love the name) a sign of what is coming or simply a gimmick?