Pattern recognition
77 views
Skip to first unread message
Bartek Kuzmicz
Aug 29, 2016, 3:53:17 AM8/29/16
to tesseract-ocr
Im writing an application that does OCR of a specific pattern that always looks as following:
AAAU123456 (4 upper case alpha characters A..Z plus 6 digits 0..9)
In addition, I have an algorithm I can use, that verifies if that entire recognized string is 100% correct - by checking the last digit, which verifies if all other characters are correct.
My question is, is there a way I can force tesseract to run in a loop (forcing it to give me alternatives) until the checking algorithm verifies it as a correct one (returns true)?
Second question is - how would you recommend I configure tesseract for that specific pattern (as described above).
BTW - the 4th character is always 'U' - the same, therefore I only need to recognize the first 3 characters plus the 6 digits.
Note; I'm using .NET - VB and/or C#
AAAU123456 (4 upper case alpha characters A..Z plus 6 digits 0..9)
In addition, I have an algorithm I can use, that verifies if that entire recognized string is 100% correct - by checking the last digit, which verifies if all other characters are correct.
My question is, is there a way I can force tesseract to run in a loop (forcing it to give me alternatives) until the checking algorithm verifies it as a correct one (returns true)?
Second question is - how would you recommend I configure tesseract for that specific pattern (as described above).
BTW - the 4th character is always 'U' - the same, therefore I only need to recognize the first 3 characters plus the 6 digits.
Note; I'm using .NET - VB and/or C#
Reply all
Reply to author
Forward
0 new messages