Hi, As you said you already tried preprocessing and it didn't work, you should train the tesseract, use the version 3 method, where training is done by individual character. version 4 uses blocks (words) doesn't usually have good results for this case.
Em seg., 7 de fev. de 2022 às 08:29, PKD <prana...@gmail.com> escreveu: > Hi, > > I have images where numbers are curved on a metal base. The standard > tesserart approach is not returning the numbers. Tried applying relevant > page segmentation method - but could not get any accurate reading of the > text. > > Any guidance \ indicator on how to approach such a problem will be very > helpful. > > Regards. > > -- > You received this message because you are subscribed to the Google Groups > "tesseract-ocr" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to tesseract-ocr+unsubscr...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/tesseract-ocr/49fda3a9-5ab3-4a70-b881-7681180351ecn%40googlegroups.com > <https://groups.google.com/d/msgid/tesseract-ocr/49fda3a9-5ab3-4a70-b881-7681180351ecn%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To unsubscribe from this group and stop receiving emails from it, send an email to tesseract-ocr+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/CAFrv86d_W6UiDjWiq1YSRozVKFOjcBEsGJ_15j5ATMhquf-oRQ%40mail.gmail.com.
