I have images like one added. img
But, Tesseract using Python have 0% accuracy (literaly). I use only 5 numbers and quality is really low. Only detecking "2" works good.
Is it about font? Someting other could be improved?
Command: d = pytesseract.image_to_string(biggest, config="--psm 10 -c tessedit_char_whitelist=12345")