Try to look at leptonica sample programs about column splitting to see if
you can preprocess the image better, before giving to tesseract


On Wed 11 Apr, 2018, 11:46 AM Ewan Mellor, <ewanmel...@gmail.com> wrote:

> Hi,
>
>
> I am using Tesseract 4 (git 10f4998a) to process a file with two columns.
> A snippet of the image is shown below.  The problem is that there is a
> fuzzy line between the two columns, and the column detector has got
> confused.  I've ended up with one block covering the first column up to
> "The" on the second line, but then a block covering both columns with the
> "patient has ..." all the way across to "history of low".
>
>
> I've looked in the debug views, and it looks to me like the line removal
> hasn't managed to remove that fuzzy line down the middle.  The "good" is
> then close enough that the column finder is deciding to merge the two
> blocks on that line.
>
>
> Looking at the code in linefind.cpp and colfind.cpp, I see lots of
> constants for various thresholds, but I don't see any configurable ones,
> and I'm not sure which way to go now.  Would it be better to work on the
> line detector in linefind.cpp and try and get rid of that vertical line?
> Or would I be better to run a columnar histogram and try and do column
> splitting myself?  Or should I ignore the fact that the line hasn't been
> removed, and concentrate on tightening up the column finder so that it's
> able to separate these two columns correctly?  It seems to me that there's
> enough of a gap there that it ought to be able to separate the columns (it
> does a pretty good job on the rest of the document, so it can't be far off).
>
>
> Any recommendations would be appreciated.
>
>
> Thanks,
>
>
> Ewan.
>
>
>
>
>
> <https://lh3.googleusercontent.com/-mrxB3T8S4fM/Ws1h25mfleI/AAAAAAAACoc/fJi8OkO6wswexnYDZU2uoofSRBCYmPiVwCLcBGAs/s1600/Screen%2BShot%2B2018-04-10%2Bat%2B6.12.48%2BPM.png>
>
> --
> 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 post to this group, send email to tesseract-ocr@googlegroups.com.
> 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/bdee5651-c305-4bbb-a14c-ccd5ba5cd7e2%40googlegroups.com
> <https://groups.google.com/d/msgid/tesseract-ocr/bdee5651-c305-4bbb-a14c-ccd5ba5cd7e2%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
> For more options, visit https://groups.google.com/d/optout.
>

-- 
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 post to this group, send email to tesseract-ocr@googlegroups.com.
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/CAG2NduWVwFi%3D-hNX_scaod%2Ba7Pp0esJmCz3MtLSAkM7PAVq%3Ddw%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to