On Wednesday, June 1, 2016 at 9:14:21 AM UTC-4, Rob Shanks wrote: > > I am trying to use Tesseract to recognise a 6 digit number from scanned > documents. Because they are scanned the numbers can be faded but I know > that they are 6 digit. The scanned documents have been destroyed long ago > so I am not able to get them rescanned. > > I have tried whitelisting 0-9 and using a user patern file \d\d\d\d\d\d > and setting psm(8) to say it is looking for one word but nothing seems to > improve detection. > > Does anyone have any suggestions? >
I'd play with modifying the images by hand to see what types of operations improve performance. The first two things I'd try would be: 1. Line removal - see if cropping/removing the thing horizontal and vertical lines improves performance 2. Dilation or other image operators to "fill in" the dropouts (ie white areas) in the numerals Tom -- 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 [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/tesseract-ocr. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/0dcfe90a-4c04-4a3f-8a9d-63ca22ee091d%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
