Infinite Path

Google may be world leader in online search but this does not mean that the company is resting on its laurels. Despite having 65% of the search market, Google has already expanded into e-mail, blogging, video, instant messaging, social networks and mobile phones. However, the company is now going back to basics by refining its search engine.

The new search revolution is called Google Goggles. It works in a similar way to a standard search engine except you use images rather than keywords. With more and more data on the web becoming pictorial, it is an obvious evolution for search. With Google Goggles, Google matches an image that has been captured on a mobile phone with those in its database, returning the most relevant information in the process. The company sees Google Goggles as a vital cog in maintaining its worldwide status and reputation as the Internet evolves into a more pictorial medium.

Aparna Chennapragada, product manager for computer vision products at Google, believes that Google Goggles is a natural progression, particularly because humans are better at seeing and recognising things than they are dealing with words, numbers and phrases.

Google Goggles is still in beta testing but has been released on Android devices running Android 1.6 and above.

Google is constantly developing new ideas, with its staff spending one day a week on product development. Another idea they have is called Image Swirl, which is similar to Google Goggles. Effectively, what Image Swirl does is take an image and analyses familiar objects within it before returning a list of alternative images which are visually similar.

It differs from a standard image search because it represents its results in the form of a spider diagram. This enables easier comparison between similar images as the results are sorted by relevance pictorially. Furthermore, one image can lead to another, which in turn can lead to another, meaning you can jump from one related image to the next in a chain. This means a map of related images from different websites is formed, which results in more relevant results, a greater library of images and sites from which to choose, and new possibilities in terms of image production and analysis.

The key to both Google Goggles and Image Swirl is recognition software, an industry which has developed rapidly in the last few years. Although both projects have yet to be fully rolled out, it is clear that pictures, images and video will form the cornerstone of the Internet in the future. And as broadband speeds increase and these forms of media become easier to view and download, Google aim to cement their place as world leader in online search.