Difficulties to use Tesseract

[Giby_the_kid]

Some years ago, I was using tesseract and was very satisfied of the
result, I tried to use it on a new computer, and it doesn't work.

[papa@localhost ~]$ tesseract out.tiff text.txt -l fra
Tesseract Open Source OCR Engine
name_to_image_type:Error:Unrecognized image type:out.tiff
IMAGE::read_header:Error:Can't read this image type:out.tiff
tesseract:Error:Read of file failed:out.tiff
Erreur de segmentation (core dumped)
[papa@localhost ~]$

Have I forgotten anything? should I install something else to make it
work properly?

OS mandriva 2010.2-64 bit, tesseract: 2.04 i586 with tesseract-fra

thanks

[Oleg Tikhonov]

Hi,
I don't know what actually you've installed, but it seems that you forgotten the libtiff.
Generally, tesseract depends on: lebtonika, and leptonika itself depends on: libz, libpbg, libtiff, libgpeg.

Make sure you have them all.

Oleg

--
You received this message because you are subscribed to the Google
Groups "tesseract-ocr" group.
To post to this group, send email to tesser...@googlegroups.com
To unsubscribe from this group, send email to
tesseract-oc...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/tesseract-ocr?hl=en

[Sriranga(78yrsold)]

Whether out.tiff is image file and if not correct name of image.tiff should be mentioned.  commandline should be "tesseract <name of image file> <output
file name> -l fra.
-sriranga(78yrs)

[zdenko podobny]

Hello,

I use tesseract on Mandrivalinux without problem. But I compiled it by myself ;-) I am not satiffied with packages provided by Mandriva team ;-) e.g. they included tesseract 3.00 to cooker but without English language data, they did not include leptonica library, that is used for image handling...)

If you are able to compile software by yourself than try the latest svn version [1]. Do not forget first install leptonica [2] ;-).

[1] http://code.google.com/p/tesseract-ocr/wiki/TesseractSvnInstallation

[2] http://leptonica.org/download.html

--

Zdenko

[Giby_the_kid]

Then after cheaking, libtiff is installed... I installed Leptonica,
but it still does not work :(

On 24 avr, 18:41, Oleg Tikhonov <olegtikho...@gmail.com> wrote:
> Hi,
> I don't know what actually you've installed, but it seems that you forgotten
> the libtiff.
> Generally, tesseract depends on: lebtonika, and leptonika itself depends on:
> libz, libpbg, libtiff, libgpeg.
>
> Make sure you have them all.
>
> Oleg
>

[zdenko podobny]

Did you recompiled tesseract? Can you send your out.tiff? (search forum for problems/limitation of tiff images)

--

Zdenko

[Will Helton]

For what it's worth, I've found that using tesseract on Ubuntu 10.10 only works if the .tiff extension is spelled with only on "f" (i.e. .tif).

If you're sure you've got everything installed (including libtiff), give that a try and see if it helps.

Regards,

Will

[Giby_the_kid]

I've test with the sample of text in the sources... it has worked...
Now if I tried with any other scanned document, I get an empty text
file.

[Quan Nguyen]

Did you scan them correctly, with appropriate pixel resolution (~300
DPI) and monochrome/grayscale settings?

[Patrick Collins]

That's weird, I find tesseract works better with 150dpi. I can never get it to return meaningful results at 300dpi. Maybe it is must my documents? Or maybe I need to force them to grayscale? They are color documents (but all black and white anyway).