Phantom Characters
30 views
Skip to first unread message
Stefan Greiner
Feb 12, 2016, 2:10:42 AM2/12/16
to tesseract-ocr
I'm using Tesseract 3.04 (953523b) with Tess4J API in a Java application.
Since the upgrade to Tesseract 3.04 (953523b) I'm getting phantom characters with the current german "deu.traineddata".
For example: The word "Marineverband" is recognised in different Versions:
-> Marine_a_verband
-> Marine_e_verband
-> Marine_verband
-> Marine_ayerband
screenshot added with source data
I get a lot of these phantom characters, I'll add different examples in the following days.
Since the upgrade to Tesseract 3.04 (953523b) I'm getting phantom characters with the current german "deu.traineddata".
For example: The word "Marineverband" is recognised in different Versions:
-> Marine_a_verband
-> Marine_e_verband
-> Marine_verband
-> Marine_ayerband
screenshot added with source data
I get a lot of these phantom characters, I'll add different examples in the following days.
Reply all
Reply to author
Forward
0 new messages