I'm trying to do a computational linguistics project, in which the methodology will create a corpus (large body of text) for analyses from the text attributed to popular memes...
I can't find any OCR capable of reading the text, due to the 'background noise' the rest of the image creates... has anyone got any ideas as to how i can get around this?
a single bump
Use a filter that makes any colour that isn't white black.
>>59523977
/thread
Simply fuck off.
>>59523851
Posting example images would be helpful.
>>59523851
Too many buzzwords. Now take a deep breath, slow down, and ask your question, this time using English words that are actually in a printed dictionary.