So, my current personal project is the ability to automatically capture screenshots from the game, OCR text and count the number of occurrences of given words.
After spending the whole evening looking around at the different OCR solutions, I realized that most OCR packages are designed for scanned text. If there are packages that can reliably read text on the screen, they are beyond this budget for amateurs.
I read some other questions, and closest I found OCR engines designed to read the screen .
It seems to me that reading the translated text should be much easier than printed and scanned text. The lines are always straight, and any given letter will always be displayed with the same image of pixels (basically, anyway). Also, why not use the actual font file (if you have one) as a cheat sheet for character recognition? We could achieve 100% accuracy with such a system.
Assuming you have a font file for cheating and the original image is completely square and has no noise, how would you recognize the characters on the screen?
(The problems I can foresee are lines and images that can confuse any rude attempt at pixel guessing.)
If you already know about OCR free / open source code for screen readers, let me know. I kind of doubt that I am going to appear, although, as it seems, none of the other seekers have received even the slightest lead.
The Python interface is preferred, but beggars cannot be selected.
EDIT:
To clarify, I'm looking for design suggestions for an OCR solution that is specifically designed to read text from screenshots. Popular tools like tesseract (mentioned in the question I linked) are hard to use at best, because they are not intended for such a source file.