Re: Issue 191 in tesseract-ocr: Doesn't recognize small text: Recognize "*354110-53153" instead of "96-020-53753", very clear image, why ?
tesser...@googlecode.com
Status: Look-here-for-help
Comment #4 on issue 191 by zde...@gmail.com: Doesn't recognize small text:
Recognize "*354110-53153" instead of "96-020-53753", very clear image, why ?
http://code.google.com/p/tesseract-ocr/issues/detail?id=191
You need to provide good input for tesseract, if you want good output.
After simple image processing (see s2.png in attachment), you can get
correct result with recent tesseract (3.01 version):
$ tesseract s2.png s2png -psm 8
Attachments:
s2.png 379 bytes
s2.txt 14 bytes
tesser...@googlecode.com
Status: WontFix
Comment #5 on issue 191 by zde...@gmail.com: Doesn't recognize small text:
Recognize "*354110-53153" instead of "96-020-53753", very clear image, why ?
http://code.google.com/p/tesseract-ocr/issues/detail?id=191
moved to FAQ:
http://code.google.com/p/tesseract-ocr/wiki/FAQ?ts=1331068532&updated=FAQ#Output_it_without_result_or_wrong