tesseract with no language words

[steven lee]

Hi,

so I have to make OCR which will scan cipher text, as we know that cipher text is not a language, so I wonder what option should I put on the language option to do this?

[Stef]

If the cipher text uses latin characters, I'd use -l eng, switch off the english word lists (load_system_dawg=0 and load_freq_dawg=0) and set the tessedit_char_whitelist to the characters that are possible in your cipher text.

Stef