jocr-devels

Re: [Jocr-devels] barcode ->pdf->email->print->fax->pdf->png->gocr

[Alessandro Z. <azu...@to...>]

On Fri, 28 Mar 2008 10:16:31 -0500
Carl Karsten <ca...@pe...> wrote:

> Anyone have any experience they can share?

 gocr will do just fine with code 128. if you want to
 use pdf417 or datamatrix, it's better to use some commercial
 software as open source ones can't keep up (yet).

 use the checksum of the barcode and/or plan
 for your own checksum.

 
-- 

 Best regards,

 Alessandro Zummo,
  Tower Technologies - Torino, Italy

  http://www.towertech.it

[Jocr-devels] barcode ->pdf->email->print->fax->pdf->png->gocr

[Carl K. <ca...@pe...>]

Don't laugh... this should work:)

Q: what free barcode font works 'best' with gocr?

Here is what I plan on doing:

database of stuff and email addresses.  email a bunch of unique pdfs, person 
prints it, signs it, faxes it to me, my fax server creates a pdf.  some process 
will create a png,  I want gocr to be able to read a barcode from the png to tie 
the fax back to the database record.

details:

At PyCon last week, I video taped every talk, including the 5 minute lightning 
talks.  The plan was to get everyone to sign a release form before they talked, 
but that didn't happen, and everyone went home.  I have a list of Talk ID, 
title, speaker name, email. (about 100.)   For each one, I will generate a PDF 
of the release form (using http://dabodev.com/wiki/ReportDesigner ), with their 
name/title filled in, email it to them.  They print it, sign it, and fax it back 
to me. (or scan/email for those that have cast aside fax.)  My fax server 
creates pdfs.   that pdf needs to get linked back to the Talk ID.   I could 
probably just print the talk ID in big human readable font that gocr could pick 
up, but some how a barcode seems better.  more sexy anyway, and that's what's 
important.

So back to the point of this post:  dabo report designer can use fonts.  I know 
of 2:

http://www.squaregear.net/fonts/free3of9.shtml
http://sourceforge.net/projects/pdf417lib

pdf417 is probably overkill, but I wouldn't mind experimenting with it, cuz I 
may be using it for some other project in 6 months or so.

Anyone have any experience they can share?

Carl K

[Jocr-devels] Handwritten character database

[Michał W. <wo...@gm...>]

Hi

I write my master thesis about handwritten character recognition and i
need some handwritten separated letters database, i have only find NIST
database but it is rather expensive(about 100$). Do you know such
database which is free or less expensive than NIST database.

Thanks for any help and sorry for offtopic, maybe when i am successful
with my master thesis i could contribute to this project.

-- 
Michał Wysokiński <wo...@gm...>

[Jocr-devels] Interested in the technique

[Anders M. <and...@gm...>]

Hi,

I have browsed your documentation, but haven't been able to grasp the
techniques you are using for the character recognition.

Do you have any pointers to the methods you are using?

Thanks in advance,

Best regards,

Anders

[Jocr-devels] GOCR Databse and Unicode Strings

[M.M. | மு.ம. <mm...@gm...>]

Dear GOCR,

I'm a newbie  to Gocr Development. I'm trying to tweak gocr to recognize My
Language, Tamil (Indic).
We are maintaining an e book project (Like Gutenberg) in Tamil, So you will
understand My needs.

Before trying ocr0.c engine, I wanted to play around with database engine.

database is working fine with Unicode code points and English chars. But
when I try to input strings both in Unicode and Latin, those strings are not
showing up in the output.
In our language, most of the characters are created by two ore three glyphs.
each glyphs has different code points. When the engine recognize a pattern,
the out put will be the combination of two ore mode code points.

How can I do this with database.

and,

even latin string are not showing up in the out put.

here is a strip from my db.lst

db_006d_1204490636.pbm 0BAE
db_006b_1204490638.pbm 00EC
db_006b_1204490639.pbm 0B95
db_006d_1204490641.pbm 0BAE
db_007a_1204490648.pbm 0BB4
db_0052_1204490659.pbm 0BB4
db_0030_1204493998.pbm 0B9F
db_006b_1204495199.pbm "0B95+0BBF"
db_0076_1204495201.pbm "0BB5+0BC1"
db_006e_1204495204.pbm 0BA9+0BCD
db_0072_1204498142.pbm 0BB0
db_0074_1204498354.pbm "thi"
db_0072_1204498356.pbm "ri"
db_0052_1204498363.pbm "Ru"
db_0074_1204498365.pbm "tha"


What;s wrong with it, and how can I use Unicode string in database in right
way?

Thanks.


-M.Mauran


-- 
http://www.mauran.blogspot.com | http://www.tamilgnu.blogspot.com |
http://www.noolaham.net | 078 514 1948 (Sri Lanka)

Re: [Jocr-devels] gOCR dictionary

[Joerg S. <Joe...@UR...>]

There is no support for dictonaries in gocr. May be in future versions.

Joerg

On Mon, 3 Mar 2008, Dario wrote:

> Hello to all
>
> I am new to the list, and excuse me for my poor English :)
>
> I would like to know if I can use a dictionary in gOCR.
> Can help me?
>
> Thank you
> Dario
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Jocr-devels mailing list
> Joc...@li...
> https://lists.sourceforge.net/lists/listinfo/jocr-devels
>
>

-- 
-------------------------------------------------------------------------
-                                                             \V/       -
- EMAIL: joe...@ur...              (o o)      -
----------------------------------------------------------oOo-(_)-oOo----
- http://www-e.uni-magdeburg.de/jschulen/
- PGP 1024D/53BDFBE3, 3816 B803 D578 F5AD 12FD  FE06 5D33 0C49 53BD FBE3
-------------------------------------------------------------------------

[Jocr-devels] gOCR dictionary

[Dario <dar...@gm...>]

Hello to all

I am new to the list, and excuse me for my poor English :)

I would like to know if I can use a dictionary in gOCR.
Can help me?

Thank you
Dario

[Jocr-devels] Another kind of character

[edbch <ed...@ya...>]

Hello to all

I am new to the list, so excuse me if I am using improperly (and especially my poor English).
I would like to look at pictures of some specific symbols, such as representing electronic components in an electronic schema.
My question is if I can use the gocr for it, and if I can, if someone could give me a roadmap than read the documentation so I can start?
Thanks for any help.

Eduardo

[Jocr-devels] Character recognition of playing cards

[Andrew L. <ala...@gm...>]

Hi All:

I am currently using GOCR v. 0.41 (I am aware that the newest version
is 0.44, I will be up grading shortly) in a playing card
identification system.  I am taking JPEG pictures, converting them to
PGM and cropping out the corners of the cards with the number and
suit.  The cropped image contains the corner of the card as well as
some of the surface the card is on (a white background).  I am using
the -m 130 option to create a database of characters to use, and it is
working quite well (~95% accuracy).  I need this to be improved
though.  It seems that if the backdrop is a solid color, it GOCR will
interpret the entire picture as a character, and it also seems that
any shadow created by the card not being flush on the backdrop can
cause errors.  I was wondering if anyone had any advice on how to
improve the accuracy of the system.  Specifically, the command I am
using is:

gocr -v 8 -m 130 -C 1234567890JQKAsdch {for suit} cropped_picture.pgm

Any help would be greatly appreciated!

Sincerely, Andrew

[Jocr-devels] How to Integrate Text Recognition with MySQL and PHP ?

[Rumah I. I. <rez...@ya...>]

Dear All,

I am newbie here.
I have my first question : How to Integrate Text Recognition with MySQL and PHP ?
I found that google do it at books.google.com. I think it can simplify books archiving on digital library than change all contents of books into text format with manual effort.

Please help me ..

Best Regards,
Reza Ervani
www.rezaervani.com


 __________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

[Jocr-devels] gocr-0.44 on OpenBSD 4.1

[<sc...@ma...>]

Hi,
I'm following the INSTALL file instructions and ran into a error.
When I run "make install" in the gocr-0.44 directory I get this error:
*** Error code 2
Stop in /root/gocr-0.44 (line 124 of Makefile).

I also tried to first run ./configure, then make and got the same error
but it stopped at line 82.

Can anyone please point me in the right direction to resolve this?

This is on OpeBSD 4.1 . There is a ports package available for OpenBSD but
it is gocr-0.41 which is not recomended to use with the FuzzyOcr plugin
for Spamassasin.

Thank you
Scott B.

[Jocr-devels] GOCR-how to handle

[Werner S. <sik...@we...>]

Hello,
I wonder if this is the right forum but I have not found any other.
I have installed GOCR on Suse10.1. When starting it, I see a little
window asking for a file. After chosing the right file ( I did it with
.jepg and with .bmp) buttons for "convert" and "view" appear. When
klicking on "view" nothing happens, klicking on"convert"  the file and
the two buttons vanish again and nothing else happens. What's going on?
Greetings,Werner.

[Jocr-devels] shape recognition

[Vlad O. <vo...@gm...>]

Hi,

I wonder if the the Jocr library
supports simple shape recognition (stars, squares, rectangles,
circles, etc). If not, is it possible to plug in your own library of
templates (stars, squares, etc) to recognize the input against? If so,
in what format?

Thanks,
Vlad

Re: [Jocr-devels] updating gtk gocr to gtk2 instead of gtk 1

[Luis M. <ga...@ot...>]

My only problem is the tests for the gtk1.2 library. I can give support
for it, if you want, of course.

Sex, 2007-09-14 às 14:05 +0200, Joerg Schulenburg escreveu:
> 
> What about removing gtk frontend completely?
> I dont have the capacity to give support for the frontend, I never tested 
> it and I dont use it. There are also no improvements the last years.
> Probably nobody use it.
> 
> Joerg.
> 
> On Tue, 4 Sep 2007, Luis Matos wrote:
> 
> > hello there
> >
> > i am trying to update the gtk 1.2 frontend to gtk2, because debian will
> > drop support for 1.2.
> >
> > but ... i am going crazy with the makefiles ... i cut off the gtk
> > ancient verification ... and i get errors that i don't know where they
> > are.
> >
> > if anybody could send me a patch to the cvs HEAD of the frontend/gnome
> > dir with this, i would apreciate ...
> >
> > oh ... and instead of gtk-config --cflags/--libs you should use
> > pkg-config gtk+-2.0 --cflags/--libs
> >
> > no gtk verification needed, for now ...
> >
> > thanks
> >
> > Luis Matos
> >
> >
> >
> >
> > -------------------------------------------------------------------------
> > This SF.net email is sponsored by: Splunk Inc.
> > Still grepping through log files to find problems?  Stop.
> > Now Search log events and configuration files using AJAX and a browser.
> > Download your FREE copy of Splunk now >>  http://get.splunk.com/
> > _______________________________________________
> > Jocr-devels mailing list
> > Joc...@li...
> > https://lists.sourceforge.net/lists/listinfo/jocr-devels
> >
> >
> 

Re: [Jocr-devels] updating gtk gocr to gtk2 instead of gtk 1

[Joerg S. <Joe...@UR...>]

What about removing gtk frontend completely?
I dont have the capacity to give support for the frontend, I never tested 
it and I dont use it. There are also no improvements the last years.
Probably nobody use it.

Joerg.

On Tue, 4 Sep 2007, Luis Matos wrote:

> hello there
>
> i am trying to update the gtk 1.2 frontend to gtk2, because debian will
> drop support for 1.2.
>
> but ... i am going crazy with the makefiles ... i cut off the gtk
> ancient verification ... and i get errors that i don't know where they
> are.
>
> if anybody could send me a patch to the cvs HEAD of the frontend/gnome
> dir with this, i would apreciate ...
>
> oh ... and instead of gtk-config --cflags/--libs you should use
> pkg-config gtk+-2.0 --cflags/--libs
>
> no gtk verification needed, for now ...
>
> thanks
>
> Luis Matos
>
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> Jocr-devels mailing list
> Joc...@li...
> https://lists.sourceforge.net/lists/listinfo/jocr-devels
>
>

-- 
-------------------------------------------------------------------------
-                                                             \V/       -
- EMAIL: joe...@ur...              (o o)      -
----------------------------------------------------------oOo-(_)-oOo----
- http://www-e.uni-magdeburg.de/jschulen/
- PGP 1024D/53BDFBE3, 3816 B803 D578 F5AD 12FD  FE06 5D33 0C49 53BD FBE3
-------------------------------------------------------------------------

[Jocr-devels] GOCR with OCR-Font

[Jan H. <ja...@ha...>]

Hi,

does GOCR works better with an OCR-Font like this?

http://www.linotype.com/de/248388/ocra-schriftfamilie.html

Thanks
Jan

[Jocr-devels] Choosing a barcode in order to improve the detection from gocr

[Marco B. <mar...@gm...>]

Hi all.

First of all, forgive me if I'm sending the email to the wrong mailing list.

Let me introduce myself. I'm one of the developer of paflow
(www.paflow.it), a program for document management inside public
administration (it is mostly targeted at italian public
administration, but it could be used also inside other european
administrations).

We're working on the optical acquisition of documents, and the
automatic association of documents to profile entered by the
employees. Let me explain this:

- a paper document arrives at the administration;

- the employee registers the document (the sender, the subject, and so
on), and prints a label with a barcode, which is sticked on the first
page of the document;

- later, all the documents are put on an high performance scanner,
which reads all the documents, produces images, which are uploaded to
paflow;

- paflow scans the images, finds the barcodes, and associates back the
images to their profiles;

Ok, I hope I've explained the enviroment. Now, I've to print a barcode
and to make the ocr (which, in this case is gocr) detect the barcode
and decode it.

I did some experiments with 3 of 9 barcode on a 99010 label, with 12
or 18 numbers encoded. The results are not very encouraging, however.

Of course, there are a lot of different parameters involved:

- the type of barcode to use (I use reportlab to generate it): 3 of 9, 128?;
- the size of the encoded data (how many numbers are encoded),
- the type of the image (i.e. tiff with/without compression, dpi, etc.)

Since I'm able to operate on many of them, I would like to have a
suggestion from those who have already experimented on it.

Regards
Marco


--
Marco Bizzarri
http://iliveinpisa.blogspot.com/
http://www.paflow.it/
http://www.icube.it/

Re: [Jocr-devels] license plate recognition

[Joerg S. <Joe...@UR...>]

On Tue, 4 Sep 2007, Janis Putrams wrote:

> Hi!
>
> I am new to this list so first I would like to say "Thank you" to all
> developers of this excellent software.
>
> My idea is to use gocr for automated license plate recognition. Has anyone
> tried anything like that?

Yes, but as I remember it was on excellent photos taken with infrared 
flash and cutted the license plate region by a seperate software.
After that gocr is fast enough.

Joerg

> I could run gocr with generic options on the pictures, but there are some
> specific things about the license plates that could optimise and speed up the
> process.
> Speed becomes very important if software triggering is used which means that
> software detects automatically when car is passing by.
>
> Thank you in advance,
> janis

[Jocr-devels] license plate recognition

[Janis P. <jan...@de...>]

Hi!

I am new to this list so first I would like to say "Thank you" to all 
developers of this excellent software.

My idea is to use gocr for automated license plate recognition. Has anyone 
tried anything like that?
I could run gocr with generic options on the pictures, but there are some 
specific things about the license plates that could optimise and speed up the 
process.
Speed becomes very important if software triggering is used which means that 
software detects automatically when car is passing by.

Thank you in advance,
janis

[Jocr-devels] updating gtk gocr to gtk2 instead of gtk 1

[Luis M. <ga...@ot...>]

hello there

i am trying to update the gtk 1.2 frontend to gtk2, because debian will
drop support for 1.2.

but ... i am going crazy with the makefiles ... i cut off the gtk
ancient verification ... and i get errors that i don't know where they
are.

if anybody could send me a patch to the cvs HEAD of the frontend/gnome
dir with this, i would apreciate ...

oh ... and instead of gtk-config --cflags/--libs you should use
pkg-config gtk+-2.0 --cflags/--libs

no gtk verification needed, for now ...

thanks

Luis Matos

Re: [Jocr-devels] greek + gocr

[Emanoil K. <del...@ya...>]

I'm very disappointed with the UTF and also greek/cyrillic support of gocr.
I would say, das-ich, forgot it - by a commercial windows app and be happy.
gocr is for me only good for playing with latin chars.

I've asked the developers to help on other character sets and the answer was that they are planing support for other languages after moving the code to use vectors.

I don't know if they moved to vectory but as far as I know the speed of development I really doubt it. (correct and forgive me if I'm wrong).

If you find an open source app that supports greek and cyrillic chars let me know.

Very sad story. Thanks for your patiens

regards


das-ich <da...@ya...> wrote: Hi all!
How do use gocr with greek character?
command gocr -i greek-text.ppm -o text -f UTF8 not use greek character, only latin.

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Jocr-devels mailing list
Joc...@li...
https://lists.sourceforge.net/lists/listinfo/jocr-devels




       
---------------------------------
Moody friends. Drama queens. Your life? Nope! - their life, your story.
 Play Sims Stories at Yahoo! Games. 

[Jocr-devels] greek + gocr

[das-ich <da...@ya...>]

Hi all!
How do use gocr with greek character?
command gocr -i greek-text.ppm -o text -f UTF8 not use greek character, only latin.

[Jocr-devels] OCR Ameatuer-Dev Debugging.

[Eric W. <ewa...@qw...>]

I am using your windows-bin (gocr044.exe) to OCR some book pages. I've 
dived in as far as I could into the "Manual" without there being a 
manual to try and figure out if I can create images better for the OCR 
to detect. In that, I've figured out how to create a db.lst, but I'm 
still experiencing minor problems in the engine I'm hoping to help 
debug. Mostly they deal with:
 
 - Font type (Times Roman I think- I'll have to go back and look)
  - interpreting w's and h's

Things I've done:
 - Made a list of all incorrectly interpreted Glyphs, and how often they 
were repeated.
 - I am currently analyzing at m -130 (not sure what it means exactly, 
but it allows gocr to create a db list)
 - Manually adjusted the levels of the pgm files to increase contrast 
and reduce dots.
 - rotated (backward correction) the image 0.55%
 - Manually configured the db.lst to adjust for problems - but still had 
issues:
    1. Engine Isolated whole and/or additional parts of letters, instead 
of one whole glyph
    2. Recieved 4 of these errors: DBG: setac(0) makes no sense!

So I do have 3 sample sets of work I can zip up. Would this be the place 
to post my attachments, or is there a more preferred way. Thanks.

  Eric

[Jocr-devels] Make example errors - resend

[Albert E. W. <ae...@AB...>]

I am resending this email due to Thunderbird losing my emails.  Please 
reply again.

I am installing the gocr-0.44 on a Mandriva 2007 Server.  I have
installed the pre-requisite netpbm-10.34.

The configure and make command both execute without an error.

Only the make example command contains the following errors.

fig2dev -L ppm -m 3 ex.fig | ppmtopcx -packed >ex.pcx
ppmtopcx: computing colormap...
ppmtopcx: 2 colors found
Error in ghostcript command
command was: gs -q -dSAFER -sDEVICE=ppmraw -r80 -g997x232 -sOutputFile=- -
*** glibc detected *** fig2dev: double free or corruption (!prev):
0x080c4048 ***
======= Backtrace: =========
/lib/i686/libc.so.6[0x401ffd9d]
/lib/i686/libc.so.6(__libc_free+0x83)[0x401fff23]
/lib/i686/libc.so.6(fclose+0x14b)[0x401ef66b]
fig2dev[0x804b531]
/lib/i686/libc.so.6(__libc_start_main+0xdc)[0x401b075c]
fig2dev[0x8049bd1]
======= Memory map: ========
08048000-080b1000 r-xp 00000000 08:06 300043     /usr/X11R6/bin/fig2dev
<snip>

My question is, Is this a signifcant failure for gocr?  Will this effect
it's use?  Or is this even limited to the use of the fig2dev tool?

TIA


-- 
Albert E. Whale, CHS CISA CISSP
Sr. Security, Network, Risk Assessment and Systems Consultant
-------------------------------------------------------------------
ABS Computer Technology, Inc. - www.ABS-CompTech.com
SPAM Zapper - No-JunkMail.com - Spam-Zapper.com - SPAM Stops Here.

[Jocr-devels] Make example errors

[Albert E. W. <ae...@AB...>]

I am installing the gocr-0.44 on a Mandriva 2007 Server.  I have 
installed the pre-requisite netpbm-10.34.

The configure and make command both execute without an error.

Only the make example command contains the following errors.

fig2dev -L ppm -m 3 ex.fig | ppmtopcx -packed >ex.pcx
ppmtopcx: computing colormap...
ppmtopcx: 2 colors found
Error in ghostcript command
command was: gs -q -dSAFER -sDEVICE=ppmraw -r80 -g997x232 -sOutputFile=- -
*** glibc detected *** fig2dev: double free or corruption (!prev): 
0x080c4048 ***
======= Backtrace: =========
/lib/i686/libc.so.6[0x401ffd9d]
/lib/i686/libc.so.6(__libc_free+0x83)[0x401fff23]
/lib/i686/libc.so.6(fclose+0x14b)[0x401ef66b]
fig2dev[0x804b531]
/lib/i686/libc.so.6(__libc_start_main+0xdc)[0x401b075c]
fig2dev[0x8049bd1]
======= Memory map: ========
08048000-080b1000 r-xp 00000000 08:06 300043     /usr/X11R6/bin/fig2dev
<snip>

My question is, Is this a signifcant failure for gocr?  Will this effect 
it's use?  Or is this even limited to the use of the fig2dev tool?

TIA


-- 
Albert E. Whale, CHS CISA CISSP
Sr. Security, Network, Risk Assessment and Systems Consultant
-------------------------------------------------------------------
ABS Computer Technology, Inc. - www.ABS-CompTech.com
SPAM Zapper - No-JunkMail.com - Spam-Zapper.com - SPAM Stops Here.