I have a huge batch of images, each one has an LED number display. I
had seen some mention in earlier discussions about someone with a
similar issue; I exclusively need to OCR numbers, no other text.
An example of what sort of image I'm talking about can be viewed here:
http://picasaweb.google.com/btreen/DigitalTextExample?feat=directlink
I've just recently learned of Tesseract and have never trained it to
recognize a new font personally, so if anyone has tackled something
like this before it'd be a huge help if you could share your files.
If not, I would also be very thankful if anyone could give me some
guidance on the "quick and dirty" method of setting Tesseract up to
recognize this sort of font (and exclusively the 10 number characters
0 - 9.)