I am basically just starting out in computer programming; mostly fluent in basic Java. I have an idea of creating an ASL (American Sign Language) to English, and my initial problem is how to identify hand movement from a webcam then comparing them to Signs that is already stored as an image or another video. If the problem is a bit too advanced for me then please list any major concepts that I can learn. Please and thank you.
Asked
Active
Viewed 434 times
-1
-
Have you heard about the gloves that can translate signed languages? http://techcrunch.com/2012/07/09/enable-talk-imagine-cup/ They work on a similar principle -- comparison to stored reference signs. Totally different detection technology, though. My understanding is that we're not there yet with visual systems. For example, the Xbox Kinect can map movements with cameras but can't see details like individual fingers. – Martin Burch Aug 04 '12 at 03:47
2 Answers
0
You clearly have a challenging problem ^^. Try to explain all you need to solve your problem would be very hard, mainly because there many ways to do this. I advice you to read a nice book about image processing (Gonzalez' book is a nice choice) and the OpenCV documentation (but it is implemented in C, C++ and has Python bindings; although it's a library that implements a lot of image processing techniques). Maybe you should focus your study on feature detection, motion analysis and object tracking. As sign language uses not just hand sign (static state) but also hand moviments (dynamic state) to express something, object tracking may be a good way to describe the signs.
I hope these informations help you, at least a little -^.^- Bye bye.

rcovre
- 69
- 4
-
Thank you very much, though I haven't actually heard of openCV means I still have a ways to go, but nonetheless I am motivated and willing/expand my skill set. – OrangesV Aug 12 '12 at 14:41
-
You're welcome. You said you know Java. Have you ever heard about JAI (Java Advanced Imaging)? Maybe it should helps you too. – rcovre Aug 12 '12 at 15:05