Searching on the Internet has always been about thinking of the right word to search by. Even the minor matter of what synonym you choose can dramatically alter the results that you get on a search. Searching for 'fat' can bring you results for anything from Fats Dominoto the FAT file system. Searching for 'obese' can give you contact information for fat farms. Google identifies ways to strategically place its AdSense ads where they are most likely to be relevant, based on words its e-mails and search queries. And entire Internet ecosystems thrive on trying to find ways to second-guess Google in finding those ways - they call it search engine optimization or SEO. And now, Google has taken its first step away from a world of keywords, into the world of key imagery, a more advanced Google search technology.
The service is called Google Goggles, and it is built for the modern Android-based camera smartphone. The basic concept is this. Think of how they have been promising you for years, a future where your computer will have three-dimensional display images, and you could just point and click on things, in the way they do on the computer screens in Minority Report, the movie by Steven Spielberg. Perhaps that kind of imagination was a little too limited for Google. Goggles, the new advanced Google search service makes the whole world your computer display. Anything you see in the world around you, a natural object, a person, a building or a business, can all be captured to be shown to Google to identify, to bring you information on it. Every object you see in the world around you might as well have a link attached, for you to click on, only, this time there is no pointed-finger over anything, a-la the computer-mouse. It is just your own.
The attempt to have computers be able to look at the world around them and identify objects is not a new one. Computer science has involved itself in this pursuit, known by the name computer vision, for ages. The problem they faced at the computer labs around the world was how they could possibly feed the computer's database with enough information about the entire world. to match images to information. They would need a computer database with millions of computers and hard disks stuffed with information ready for super fast retrieval. Well what do you know, this is exactly the kind of database Google has built for its search engine business over the last decade in its worldwide network of server farms and data centers. The technology to recognize images of objects is not Google's innovation. Providing computing power and meticulously organized data, pulled together with intuitive search, is the advanced Google search contribution. Indeed, there have been other applications that try to allow you to do this. World Surfer and Wikitudefor example. But these rely entirely on GPS to try to find something. But they use satellites to locate the place you are, and try to guess what you're pointing your camera at. Goggles does that too, but its image identification technology and the huge repository of images it has, take things to a different level.
Advanced Google search technologies will take a picture you snap, and distribute it to all its data centers across the world instantly, to match it against millions of images from its databases. The strategy of distributing workload, brings back results nearly instantly. Certainly, this is a nascent technology; recognizing people for instance, even people on Google's database, can be quite iffy. However it does recognize the businesses you might be feeding it pictures of, very well. If you shoot the front of a McDonald's outlet for though, it will have problems - McDonald's looks the same everywhere. If you happen to be looking at a great-looking plasma TV at an airport, and want to know the model number, or if you are passing through a strange part of town and don't know whom to ask for directions, then Goggles will be just the ticket when it comes out of beta .
0 Response to "Advanced Google Search Technology - Search and Ye Shall "
Post a Comment