I'm trying to OCR photos of text on a PC screen. I'd think OCR would do a much better job if I could give it hints, such as:

• the text to recognize is in a monospaced font (now, it tries to use a font with variable character widths, frequently leading to words with spaces in the middle of them)

• the text to recognize has regular line spacing

Also, it would do much better if I could tell it:

• proceed line by line; don't jump all over the place. OCR sometime reorders the content of tables in strange ways.

Attached is an example where OCR did a decent, but not great job. With hints, I'd think it could approach 100% accuracy.

Thanks.