Hi all!
It is very interesting for me if someone had any tests with different CPU, memorry e.t.c
Is it real to make tesseract recognize a difficult image(small font, tables, dirty image) less then 1 second.
If real then what I need to use for the best result?