Hi all,
you will not get OSD (Orientation and script detection) output information
with tesseract executable. At the moment tesseract provide (save) only ocr
result. Somebody could consider help (tesseract --help) misleading because
it enumerate all possible page segmentation modes. I think that starting
with other than 0 psm would also trigger question...
OSD info you get via API. Here is simple snippet:
inputfile = "/usr/src/tesseract-3.02/eurotextUpsideDown.png";
image = pixRead(inputfile);
api->Init("/usr/src/tesseract-3.02/", "eng");
api->SetPageSegMode(tesseract::PSM_AUTO_OSD);
api->SetImage(image);
api->Recognize(0);
tesseract::PageIterator* it = api->AnalyseLayout();
tesseract::Orientation orientation;
tesseract::WritingDirection direction;
tesseract::TextlineOrder order;
float deskew_angle;
it->Orientation(&orientation, &direction, &order, &deskew_angle);
printf("Orientation: %d;\nWritingDirection: %d\nTextlineOrder: %d\n" \
"Deskew angle: %.4f\n",
orientation, direction, order, deskew_angle);
In output you can see these information(eurotextUpsideDown.png is 180
degree rotated image eurotext.tif):
Orientation: 2;
WritingDirection: 0
TextlineOrder: 2
Deskew angle: -0.0038
This means that page is upside down (Orientation)[1], text is written from
left to right (WritingDirection)[2], lines are ordered from top to down
(TextlineOrder)[3], and there is small deskew angle[4].
As you can see attributes WritingDirection, TextlineOrder, Deskew do not
reflect that page is upside down, so you will get the same result for
eurotext.tif.
[1]
http://code.google.com/p/tesseract-ocr/source/browse/trunk/ccstruct/publictypes.h?r=716#81
[2]
http://code.google.com/p/tesseract-ocr/source/browse/trunk/ccstruct/publictypes.h?r=716#111
[3]
http://code.google.com/p/tesseract-ocr/source/browse/trunk/ccstruct/publictypes.h?r=716#125
[4]
http://code.google.com/p/tesseract-ocr/source/browse/trunk/ccmain/pageiterator.h?r=716#239
--
Zdenko
On Tue, Nov 13, 2012 at 10:26 AM, Alex <[email protected]> wrote:
> Hello,Chirag.
>
> I am also trying to find a way to detect the script of the input document.
>
> Kindly let me know if you have some progress.
>
> Thanks and Regards,
> Alex
>
>
>
> 在 2012年3月14日星期三UTC+8下午7时17分29秒，Chirag Jain写道：
>>
>> With -psm 3, I got non-empty files (test_osd.txt) which were empty with
>> -psm 0. This is true for both with/without -l options.
>>
>> However, the results of detectOS is same for both -psm [0/3] option for
>> any of with/without -l options.
>>
>> Please note that I have modified the code slightly to call detectOS
>> separately, which has been doing a good job for orientation detection given
>> script. I am struggling to detect the script of the input document.
>>
>> Regards,
>> Chirag
>>
>>
>> On Wed, Mar 14, 2012 at 4:05 PM, Sriranga(78yrsold)
>> <[email protected]>wrote:
>>
>>> one more important - please test again as follows:
>>> 1st test:tesseract.exe japanese_doc.tif test_osd -l jpn -psm 3
>>> 2nd test:tesseract.exe japanese_doc.tif test_osd -psm 3
>>> Please check the output text files "test_osd" - you will find
>>> difference in script between two.
>>>
>>> On Wed, Mar 14, 2012 at 3:51 PM, Sriranga(78yrsold) <[email protected]
>>> > wrote:
>>>
>>>> I noticed "-l lang" before "-psm 0" is missing in your commandline.
>>>> In the absence of "-l lang" tesseract will always assume as "-l eng".
>>>>
>>>> extract of help is reproduced below:
>>>>
>>>> M:\>tesseract.exe -h
>>>> *Usage:tesseract.exe imagename outputbase [-l lang] [-psm pagesegmode]
>>>> [configfil*
>>>> e...]
>>>> pagesegmode values are:
>>>> 0 = Orientation and script detection (OSD) only.
>>>> 1 = Automatic page segmentation with OSD.
>>>> 2 = Automatic page segmentation, but no OSD, or OCR
>>>> 3 = Fully automatic page segmentation, but no OSD. (Default)
>>>> 4 = Assume a single column of text of variable sizes.
>>>> 5 = Assume a single uniform block of vertically aligned text.
>>>> 6 = Assume a single uniform block of text.
>>>> 7 = Treat the image as a single text line.
>>>> 8 = Treat the image as a single word.
>>>> 9 = Treat the image as a single word in a circle.
>>>> 10 = Treat the image as a single character.
>>>> -l lang and/or -psm pagesegmode must occur before anyconfigfile.
>>>>
>>>>
>>>>
>>>> On Wed, Mar 14, 2012 at 3:22 PM, Chirag <[email protected]> wrote:
>>>>
>>>>> Hi all,
>>>>>
>>>>> I was able to successfully test orientation detection (after stepping
>>>>> though the code) for various scripts using following commands:
>>>>>
>>>>> English: tesseract.exe english_doc.tif test_osd -l eng -psm 0
>>>>> Japanese: tesseract.exe japanese_doc.tif test_osd -l jpn -psm 0
>>>>> Korean: tesseract.exe korean_doc.tif test_osd -l kor -psm 0
>>>>>
>>>>> In these cases, the executable search for eng.traineddata,
>>>>> jpn.traineddata and kor.traineddata respectively along with
>>>>> osd.traineddata.
>>>>>
>>>>> The performance is really good.
>>>>>
>>>>>
>>>>> However, it seems like Tesseract is detecting orientation given script.
>>>>>
>>>>>
>>>>> If I run the executable as following:
>>>>>
>>>>> Japanese: tesseract.exe japanese_doc.tif test_osd -psm 0
>>>>> Korean: tesseract.exe korean_doc.tif test_osd -psm 0
>>>>>
>>>>> The results are not good. It seems like script detection is not robust.
>>>>>
>>>>> Am I missing some step? Kindly clarify.
>>>>>
>>>>>
>>>>> Regards,
>>>>> Chirag
>>>>>
>>>>>
>>>>> On Sat, Mar 3, 2012 at 7:12 PM, koray <[email protected]>wrote:
>>>>>
>>>>>> OSD returns emty text when I tried. Can anyone please clarify if
>>>>>> this is a bug or I m doing things wrong?
>>>>>>
>>>>>> --
>>>>>> You received this message because you are subscribed to the Google
>>>>>> Groups "tesseract-ocr" group.
>>>>>> To post to this group, send email to [email protected]
>>>>>>
>>>>>> 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<http://groups.google.com/group/tesseract-ocr?hl=en>
>>>>>>
>>>>>
>>>>>
--
You received this message because you are subscribed to the Google
Groups "tesseract-ocr" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/tesseract-ocr?hl=en
