http://code.google.com/p/tesseract-ocr/wiki/TrainingTesseract3
i did training by reading this document.
2011/10/28 Sven Pedersen
> Hi Bui Van,
> Have you read the documents on the website? What font do you need?
> --Sven
>
>
> On Thu, Oct 27, 2011 at 9:15 PM, bui van Chuong > wrote:
>
>> Hello
few months' archives. Perhaps
> you should post the question to StackOverflow.com?
> --Sven
>
>
>
> On Thu, Oct 27, 2011 at 9:12 AM, merve t wrote:
>
>> no idea?
>>
>> 2011/10/24 merve t
>>
>>> Hello,
>>>
>>> First of all i
no idea?
2011/10/24 merve t
> Hello,
>
> First of all i want to confirm that i understand choice iterator right.
>
> For example if i have a word on an image "scope",
>
> choice iterator must give me something like "s" and maybe after Next(),
> &qu
give a
> result.
>
> Thanks for the answers.
>
> On Oct 25, 7:57 am, merve t wrote:
> > Yes, can you explain PSM 8?
> > Is it something like PSM_AUTO, should i change pagesegmode to PSM_8?
> > In this mail group, i was advised to give tesseract characters one by
&g
Yes, can you explain PSM 8?
Is it something like PSM_AUTO, should i change pagesegmode to PSM_8?
In this mail group, i was advised to give tesseract characters one by one.
Thus i must learn how to make tesseract recognize alone chars in images.
Thanks in advance.
2011/10/24 patrickq
> What's PSM
).GetUTF8Text(tesseract::RIL_SYMBOL);
printf("out:%s,out2:%s",out,out2);
output on console is:
(null),P
p is expected result iterator result, but choice iterator output is null.
thanks for ideas.
2011/10/18 merve t
> I tried it too, but could not succeed to get choice iterations
;depth/8,img->widthStep);
i think problem is here, but do not know exactly where.
2011/10/22 merve t
> Hello,
>
> I use pageIterator's boundingbox function like this:
>
>
> int left,top,right,bottom;
> left=0;top=0;right=0;bottom=0;
> api.Recognize(NULL);
>
Hello,
I use pageIterator's boundingbox function like this:
int left,top,right,bottom;
left=0;top=0;right=0;bottom=0;
api.Recognize(NULL);
tesseract::ResultIterator *ri=api.GetIterator();
char * sonuc=(*ri).GetUTF8Text(tesseract::RIL_SYMBOL);
if((*ri).BoundingBox(tesseract::RIL_WORD,&left,&top,&
Hello,
i need to provide a pix(of a char) to tesseract and hope it recognizes.
To try if i achieve this i did some internal experiments.
writing these code lines
Pix* pixa=(*ri).GetBinaryImage(tesseract::RIL_SYMBOL);
api.SetImage(pixa);
char *sonuc=api->GetUTF8Text();
printf("%s",sonuc);
i hop
hin ScanBizCards we
> question and test every space (as well as insert spaces where we feel
> one is missing). We are having decent success but still have a lot to
> improve there.
>
> Patrick
>
> On Oct 18, 5:29 am, merve t wrote:
> > Hello,
> > when i write "apple&qu
Hello,
when i write "apple" into a file tesseract some times resolve it as "app
le".
Do i have the chance to specify some value in some where and make tesseract
resolve some letters as they are in the same word?
Thanks
--
You received this message because you are subscribed to the Google
Groups "
seract may be
> using 8-bit/pixel images, and that may be what's being returned from
> the GetBinaryImage() - the code you've presented appears to expect >=
> 24bit/pixel image...
>
> On Oct 17, 9:39 a
I tried it too, but could not succeed to get choice iterations. I am going
to try exact solution by that topic's itarator. If i succeed, i am going to
write here too.
2011/10/17 Stane
> Hi,
>
> i used here in this thread
>
> http://groups.google.com/group/tesseract-ocr/browse_thread/thread/82125
Hello,
I wrote this code:
tesseract::ResultIterator* ri=api.GetIterator();
Pix* pixa=(*ri).GetBinaryImage(tesseract::RIL_SYMBOL);
l_uint32 width=pixGetWidth(pixa);
l_uint32 wpl=pixGetWpl(pixa);
l_uint32 h=pixGetHeight
Or there is a way in tesseract, for example a method which is not protected?
One point; i am aware of existence of pageiterator::getimage but i do not
want to get a symbol, i want to get independent black clusters.
Thanks for any idea.
--
You received this message because you are subscribed to th
Hello,
Is there somebody have ever used Choice Iterator?
I can not achied to use it successfully.
Thanks for any idea.
--
You received this message because you are subscribed to the Google
Groups "tesseract-ocr" group.
To post to this group, send email to tesseract-ocr@googlegroups.com
To unsubsc
or do you advise to use an other librery like opengl to get char images?
Thanks for time.
2011/10/5 merve t
> Hello,
> Thanks for your encouraging response, i was about to give up to try.
>
> I am using PSM_AUTO
>
> What is the most elegant way of getting images char by ch
hen hand
> over every character to Tesseract for recognition in PSM_CHAR mode..that
> should solve the matter...which mode are you using Tesseract in currently?
>
> --
> Regards,
> Saurabh Gandhi
>
>
>
>
>
> On Tue, Oct 4, 2011 at 8:03 PM, merve t wrote:
>
>>
Hello,
I think my problem is related to this.
But a little different
for example in my image there is
"lt"
and tesseract recognizes this as "H".
i do not want tesseract to combine blobs, i want it recognize blobs
individually.
What should i do?
2011/10/4 Saurabh Gandhi
> Hi Rakesh,
>
> We d
Thanks very much, i found them in usr/local/include/tesseract
2011/9/30 zdenko podobny
>
>
> On Fri, Sep 30, 2011 at 6:20 PM, merve t wrote:
>
>> Hello,
>> I installed tesseract-ocr software on ubuntu linux by downloading from svn
>> and compiling and making and i
How should i use ChoiceIterator?
I want to get all word's and possibility confidences, for example
if in my text it wrotes
merve
and if tesseract have 3 results like
m€rv€
metve
met7e
i want to know all of them.
I tried so far:
tesseract::ResultIterator* ri=api.GetIterator();
tesseract::Cho
ing
> > > the trained file?
> >
> > > Which version are you using ?
> >
> > > Did you try to use your (successful) trained file ?
> >
> > > I have bumped into an old problem like yours on
> > >http://code.google.com/p/tesseract-ocr/issues/det
Hello,
I installed tesseract-ocr software on ubuntu linux by downloading from svn
and compiling and making and installing.
Now i am looking at /usr/include directory and can not see tesseract related
headers.
Also when i try to compile a source file of mine which includes headers of
tesseract, i g
nto an old problem like yours on
> http://code.google.com/p/tesseract-ocr/issues/detail?id=490. Official
> poster there suggested removing the first character from the box file
> or the last character, which avoids the crash.
>
> Hope it helps.
>
> Cihan
>
> On Sep 29, 4:
Hello,
a few days ago i was be able to train tesseract but yesterday and today i am
getting the error:
Reading zxc.denemem.exp21.tr ...
Error: Illegal feature parameter spec!
Fatal error: No error trap defined!
Signal_termination_handler called with signal 1000
My box file is seems to be corre
Hello,
Is there somebody have ever used ResultIterator?
I do not know if i understand right but i need all possibilities for one
word in the text.
There is a pageIteratorLevel property, i think i must pass all paragraph,
line, word and symbol number to result iterator to get text and confidence.
Hello,
I know you are about to kick me from the list for too much asking but
can i use trainneddata in Linux, even if i generate the trainneddata in
Windows?
--
You received this message because you are subscribed to the Google
Groups "tesseract-ocr" group.
To post to this group, send email to t
Hello,
I am trying to train Tesseract for handwriting(for now with not adjacent
letters) and get good(%60 success) results.
My main problem now;
"rin" is recognized as "m"
i do not want change unicharambigs because, some time m can be truely
recognized.
I want to ask can i get the possibilities
Hello,
I am trying to train tesseract
and i DO NOT beleive that no one knows the meaning of
blk
row
and allrows
in the error lines:
APPLY_BOXES: Unlabelled word blk:1 row:5 allrows:5
APPLY_BOXES: Unlabelled word blk:1 row:5 allrows:5
what is the coordinates, my training data contains utf8
Hello,
I am trying to train Tesseract for Turkish and there is a letter and there
is a problem with uppercase Ğ, if you can not see upper case G and a its hat
on it. There is no error about lower case g with its hat.
I make box file and edit it
then run this command:
C:\Program Files\Tesseract-
Thanks
2011/9/21 Giuseppe Menga
> Pix and Box are two data structures of Leptonica containing an image and
> a box (4 dimensions), respectively..
> Pixa and Boxa are arrays of Pix and of Box.
> See documentation of Leptonica.
> Giuseppe
>
> *From:* merve t
> *Sent:*
Hello,
I am trying to look at all the code to find something.
I saw a line:
Boxa* TessBaseAPI::GetRegions(Pixa** pixa) {
i searched in the doc's index.html file, but there is no class or structure
named Boxa and Pixa.
Thanks
--
You received this message because you are subscribed to the Googl
g into Tesseract's code.
>
> Warm regards,
> Dmitri Silaev
> www.CustomOCR.com
>
>
>
>
>
> On Mon, Sep 19, 2011 at 9:55 PM, merve t wrote:
> > Hello,
> > or can i get it as a image with a transparent background?
> >
> > I think tesseract f
Hello,
I am simplifying my question.
What is the name of the class and method that returns blobs(i assume blob is
the shape of a letter or a word which consists of contigious black pixels)
2011/9/19 merve t
> Hello,
> or can i get it as a image with a transparent background?
>
Hello,
or can i get it as a image with a transparent background?
I think tesseract first detects the pixels of the letters, it gets the shape
of the different colored pixels on image while solving images.
I do not know if i am true,
but if i am;
i want to use tesseract to get the drawn shape's
priate results.
For the attached file, i want to get 4 letters "in testing step", it does
not matter if the first letter is "d" or "g" or "a".
But i get 6 letters "fcgbcd".
I can not understand it.
2011/9/16 merve t
> Hello,
&
Hello,
I trained tesseract with a little dataset of my hand writing and i get some
good results, but when i try to "test" the image attached i get
"fcgbcd"
as output.
the last three chars are correct "bcd".
But for "a" it returns "fcg" , three chars.
As another process i tried to generate a bo
> Offset for type 6 is -1
> Offset for type 7 is -1
> Offset for type 8 is -1
> Offset for type 9 is -1
> Offset for type 10 is -1
> Offset for type 11 is -1
> Offset for type 12 is -1
>
> M:\>tesseract example.png testexample -l han
> Tesseract Open Source OCR Engine wi
i made everythin g carefully now it working
thanks
2011/9/16 zdenko podobny
> Yes, you are right - combine_tessdata do not make a bnm.trainneddata in
> tessdata folder. It never promiss to do it.
>
> If you need reply with more details, first you must povide details :-) E.g.
> OS, tesseract/comb
gt; It is possible in tesseract -vide attached files which is self explanatory.
> Cheers,
> -sriranga(78yrs)
>
> On Mon, Sep 12, 2011 at 6:04 PM, merve t wrote:
>
>> Hello,
>> There is a file attached,
>> I must confess, i wrote it with mouse, but the data that
Hello,
There is a file attached,
I must confess, i wrote it with mouse, but the data that is needed to be
solved is like this.
Because we are developing a white board application.
I tried to solve it with ocropus but it could not.
I can not install tesseract alone, if you say it can solve this pic,
CellWriter needs to be trained but i need a handwriting recognition which
can recognize many different writers. Do you know about CellWriter, can i
use it by training it with a universal data set so it can recognize
different writers without training for every writer?
Maybe it is not the correct pl
Then, i can not use it because i want to recognize freely written
handwritings.
Thanks for reply.
2011/8/24 Dmitri Silaev
> Simple answer: in general, no.
> However, in particular, it might.
> Send sample images to get a more certain answer
>
> Warm regards,
> Dmitri Silaev
> www.CustomOCR.com
>
43 matches
Mail list logo