The most simple use of tesseract in C++ app

[braza]

Hi,

The world of open source welcomes me with insufficient info/examples/
documentation but with opened doors to ask ;)

I`m trying just to recognize really clear and simple line of text in
English like "Tess TEST 123.4 $15"

now I have:

//Tesseract block start
CTessOCR *tess = CTessOCR::Instance();
tess->api->SetVariable ("tessedit_char_whitelist", "0123456789");
tess->api->SetVariable ("classify_bln_numeric_mode", "1");
tess->api->Init ("./../../tessdata", "eng");
#ifdef DEBUG_MODE
tess->api->SetVariable ("tessedit_adaption_debug", "T");
tess->api->SetVariable ("tessedit_draw_outwords", "T");
tess->api->SetVariable ("tessedit_dump_choices", "T");
tess->api->SetVariable ("tessedit_dump_choices", "T");
tess->api->SetVariable ("interactive_mode", "T");
tess->api->SetVariable ("tessedit_create_hocr", "T");
#endif
tess->api->SetVariable ("tessedit_single_match", "0");
tess->api->SetVariable ("tessedit_zero_rejection", "T");
tess->api->SetVariable ("tessedit_minimal_rejection", "F");
tess->api->SetVariable ("tessedit_write_rep_codes", "F");
tess->api->SetVariable ("tessedit_resegment_from_boxes", "T");
tess->api->SetVariable ("tessedit_train_from_boxes", "T");
tess->api->SetVariable ("textord_fast_pitch_test", "T");
tess->api->SetVariable ("textord_no_rejects", "T");
tess->api->SetVariable ("edges_children_fix", "F");
tess->api->SetVariable ("edges_childarea", "0.65");
tess->api->SetVariable ("edges_boxarea", "0.9");
tess->api->SetVariable ("il1_adaption_test", "1");
tess->api->SetPageSegMode (tesseract::PSM_SINGLE_LINE);

Mat img = imread( "../../tess.jpg", CV_LOAD_IMAGE_GRAYSCALE ); //err
now
tess->api->SetImage(convert_mat_to_pix(img));
std::string text = tess->api->GetUTF8Text();

It all fails in

> match.exe!OpenBoxFile(const STRING & fname)
match.exe!tesseract::Tesseract::ApplyBoxes(const STRING & fname,
bool find_segmentation, BLOCK_LIST * block_list)
match.exe!tesseract::TessBaseAPI::Recognize(ETEXT_DESC * monitor)
match.exe!tesseract::TessBaseAPI::GetUTF8Text()

Obviously it fails because I`ve never set input file name with boxes.
But why would I need it? I already have trained data downloaded and
put in tessdata: eng.traineddata, eng.cube.size etc

[zdenko podobny]

You can try this:

$ convert -rotate 10 phototest.tif phototest-r.png

$ g++ -o test test.cpp -I/usr/local/include/tesseract/ -I/usr/local/include/leptonica/ -L/usr/local/libs -ltesseract

$ ./test

where:

phototest.tif  is from tesseract source

convert - is part of imagemagick. First line is not needed ;-)

Tested on Mandrivalinux with tesseract 3.01

----------------  test.cpp ----------------

#include <baseapi.h>

#include <allheaders.h>

Pix *pix;

int offset;

float slope;

int main(int argc, char *argv[]) {

    tesseract::TessBaseAPI api;

    

    pix = pixRead("phototest-r.png");

    

    api.Init(NULL, "eng");

    api.SetPageSegMode(tesseract::PSM_AUTO); 

    api.SetImage(pix);

    

    char* outText = api.GetUTF8Text();

    int conf = api.MeanTextConf();

    api.GetTextDirection(&offset, &slope);

    

    printf("Confidence=%d\n", conf);

    printf("Offset: %d Slope: %.2f\n", offset, slope);

    printf("OCR output:\n\n");

    printf(outText);

    api.Clear();

    api.End();

    delete [] outText;

    pixDestroy(&pix);    

    return 0;

}

----------------  test.cpp ---------------- 

I hope this help a little bit.

Zdenko

--
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

[braza]

Many thanks! It worked, so at least I can move forward from your
scratch.

On 18 дек 2011, 02:03, zdenko podobny <zde...@gmail.com> wrote:
> You can try this:
>
> $ convert -rotate 10 phototest.tif phototest-r.png
> $ g++ -o test test.cpp -I/usr/local/include/tesseract/
> -I/usr/local/include/leptonica/ -L/usr/local/libs -ltesseract
> $ ./test
>
> where:

> *phototest.tif*  is from tesseract source
> *convert* - is part of imagemagick. First line is not needed ;-)

[zdenko podobny]

it looks like you need to specify leptonica library for linking (-llept)

Dňa 14.6.2012 9:13, "pcollinse" <patrick...@vsig.com> napísal(-a):

Thanks for this wonderfully simple example!

It compiles for me on Ubuntu, but not on Mac OSX. Anyone know what I'm missing here?

$ g++ -o test test.cpp -I /opt/local/include/tesseract -I /opt/local/include/leptonica -L /opt/local/lib -ltesseract

Undefined symbols for architecture x86_64:

  "_pixRead", referenced from:

      _main in ccVNVpU7.o

  "_pixDestroy", referenced from:

      _main in ccVNVpU7.o

ld: symbol(s) not found for architecture x86_64

collect2: ld returned 1 exit status

Patrick.

[Dovhani Foneworx]

Good day; I just wanted to ask if I can do the training using the library as like the example you are showing in this example.

My problem is I don't see a clear explanatory which function to use when I want to do for example "mftraining" command from my C code instead of using the system command.