Real-time translation was announced for Google's mobile application Google Goggles today. Previously, the app already used the camera on a handset that runs the Android mobile operating software to directly search the Web for mor
e information on the photo subject. The user takes a picture, the service uploads it to the server, and the object or place is matched with relevant material. Now this same image, if it contains text, will trigger Google Goggles to respond with translated text.
Translation joins the suggested list of different subjects that Google Goggles can return useful information on: landmarks, artwork, places, logos and barcodes. According to the Google Mobile blog, at its prototype release in Barcelona at the Mobile World Congress, it could only recognize German text, but can now read "English, French, Italian, German and Spanish and can translate to many more languages." Work is currently underway to expand to non-Latin script based languages, a matter that will only become more relevant with Web domain addresses expanding to Arabic and more scripts.
Seen as its first move into actual utility, Engadget referred to GG as only good for "telling you that your can of Coke looks like someone's face." Other improvements the blog identified in this update were better barcode and image recognition, an improved user interface, and the ability to identify material from within a user's image gallery. But the translation functionality is what is stirring up responses - any traveler can imagine how useful this could be in a country with an unfamiliar language. The introduction video highlights this by translating an entry on a German menu, but could easily be applied to transit stations, street signs, and more.
For phones that run Android 1.6 and higher, Google Goggles is available at the Android Market or by scanning the code at the Mobile blog article.