How to make Augmented Reality to translate language in Android ?

Processing IDE could help.Its a graphic sketch IDE.but libraries can imported.Image processing is pretty easy using this IDE.Export as Android api is provided.

Can you explain more please :)

what kind of video translator ?!

You can go for unity 3d or opencv...

If you want to make a sound translation, it is case for sound processing. Augmented reality (or image processing) is in this case most probably out of question. Maybe just in case of the sign language or non-verbal communication. In that case you will need most probably OpenCV or similar library.

Video-translator translates the unknown language into native language using augmented reality in real time.(Aditya pandu pls rectify me if that is not wat u r looking for).OpenCV wrapper and processing could get ur work done faster(in real time).Matlab would not do it in real time, it would be sluggish too.

Just to be sure I understand the gist: You want to create an mobile application that could be pointed on a person. On the screen will be face of the speaking person and the audio stream will be translated into some language. If this is the application, IMHO there is no need for image processing. From im. proc. you could get just the words that are allready (in a clear form) in the audio stream. The application therefore just presents the camera stream (this could be done using OpenCV, but also by many other libs) a the question is how to translate the sound in the way the Google does in their translator (http://techcrunch.com/2009/12/14/the-unofficial-google-text-to-speech-api/, http://mikepultz.com/2011/03/accessing-google-speech-api-chrome-11/)...

To All : Video translator that I mean like this video : http://www.google.co.id/url?sa=t&rct=j&q=augmented%20reality%20video%20tranlsate&source=web&cd=1&ved=0CCAQtwIwAA&url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dh2OfQdYrHRs&ei=YQQqT8zNLIfMrQfSw4HhDA&usg=AFQjCNHzmkFb6yY1jb6J0S1KMjicCmtydw&cad=rja

But I want to make it in Android,.. Can you help me ??? Please,..

That is quite "simple" ;-) First: You need a good OCR application. You could use some projects from Google Code. However, these free API for web services or free OCR libraries are not perfect (in fact for my project they were useless). You could also buy a professional OCR library. Nevetherless, that is not cheap. The last option is to use unofficial API of Google Googles (browse the net for examples). This last option is not stable (it is not an official API), but it is really well working.

When you get a text, you just need to translate it using some on-line service (again Google is not bad...).

Workshop on Digital Image Watermarking

Start with a concept. Ask yourself, what it is the user does, what he sees, hears etc. and how that is used to trigger a response. That way your concept will become more clear and you see what is new or even unique.

If you say it is like in that video... well, there it is already.

hmm ... from your answers and comments seem to have deep technical skills. but it seems not that easy, just by way of OCR, with unity 3D (because it's only for 3D modeling in AR) and has nothing to do with Digital Image Watermarking (no correlation in this case). In AR there are have conditions in perspective image and augmentation object in order to follow the conditions of perspective, it can only be solved by marker detection algorithm (marker-based type), or SURF/SIFT (markerless type) and then use perspective projection, after that you can use OCR technique to read text inside marker. for detecting marker there are already framework to do that on android platform, the framework is Qualcomm vuforia

in fact many AR framework besides vuforia that can be run on android platform, like ARToolKitPlus, NyARToolKit, ARTag, StudierStube, MetaIO, DFusion and many more...

hey Pandu, you are from Indonesia, right....?

Indonesia has a facebook group that focuses on AR or maybe you already know, the group is

"ART (Augmented Reality Team)" by Anggie from Lampung University and "Augmented Reality Forum Indonesia (ARFI)" by AR & Co Indonesia (I have worked there), you can discuss there, the group that is more easily understood, tutorials and practical but less theoretical

i want to make application like this.. but it's used for "papan informasi petunjuk jalan" (Direction board or road sign) so it will translate only the text in within.. (the translation from indonenesian-english)

i agree that in order to make the app have to use OCR and AR SDK.

You have to understand some methods to achieve augmented reality, so you can insert OCR methods and perform language translation.

The most important issue in AR marker-based/markerless is perspective distortion, if understood, it will easily insert other methods