[Bug-ocrad] Questions about GNU OCR Capabilities

[Dylan Klett]

Hi,
I'm a web developer trying to build a Rails app that includes OCR from a
mobile-first angle. Most OCR I've run into so far seems good at recognizing
text, but I'm also looking for it to recognize pictures, symbols, etc.
Would GNU OCR be able to do this? Or is there a "learning" feature or
similar where I could add pictures to its library of recognizable
text/images/characters?

I'm linking an example of the kinds of pictures/symbols I'd want the OCR to
be able to (or learn to) recognize. I'd like to be able to have the OCR
recognize the four symbols individually, then assign a value to them
individually to be called on later. Thanks!

Thanks for the help!
Dylan

Link to images:
http://i.imgur.com/P6sWMZm.png