Combining tessdata files Error opening unicharset file

[Miguel Goyanes]

Hello.

I'm reading the tutorial on hoe to create trainedata and I came across the error in the title jsut at the last part.

I'm running Ubuntu 14.04 LTS on a Virtual machine.

Tesseract version is 3.02.02 abd leptonica is 1.72

Hi have a por.monospaced.exp0.tif file and here are the steps I've made and resulting output

miguel@miguel:~/Desktop/TessTests$ tesseract por.monospaced.exp0.tif por.monospaced.exp0 batch.nochop makebox
Tesseract Open Source OCR Engine v3.02.02 with Leptonica

Then:

miguel@miguel:~/Desktop/TessTests$ tesseract por.monospaced.exp0.tif por.monospaced.exp0 box.train

Tesseract Open Source OCR Engine v3.02.02 with Leptonica

APPLY_BOXES:

   Boxes read from boxfile:      59

   Found 59 good blobs.

   Leaving 2 unlabelled blobs in 0 words.

TRAINING ... Font name = monospaced

Generated training data for 4 words

And  

miguel@miguel:~/Desktop/TessTests$ unicharset_extractor por.monospaced.exp0.box

Extracting unicharset from por.monospaced.exp0.box

Wrote unicharset file ./unicharset.

I've created the font_properties file

 

miguel@miguel:~/Desktop/TessTests$ echo monospaced 0 0 0 0 0 > font_properties

And

miguel@miguel:~/Desktop/TessTests$ shapeclustering -F font_properties -U unicharset por.monospaced.exp0.tr

Reading por.monospaced.exp0.tr ...

Building master shape table

Computing shape distances...

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0

Stopped with 0 merged, min dist 999.000000

Computing shape distances...

Stopped with 0 merged, min dist 999.000000

Computing shape distances...

Stopped with 0 merged, min dist 999.000000

Computing shape distances... 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Distance = 0.000000: Stopped with 1 merged, min dist 0.165217

Master shape_table:Number of shapes = 20 max unichars = 2 number with multiple unichars = 1

miguel@miguel:~/Desktop/TessTests$ mftraining -F font_properties -U unicharset -O por.unicharset por.monospaced.exp0.tr

Read shape table shapetable of 20 shapes

Reading por.monospaced.exp0.tr ...

Done!

And then

miguel@miguel:~/Desktop/TessTests$ cntraining por.monospaced.exp0.tr

Reading por.monospaced.exp0.tr ...

Clustering ...

Writing normproto ...

Finally,in the last command I'm getting the error:

miguel@miguel:~/Desktop/TessTests$ combine_tessdata pass.

Combining tessdata files

Error opening unicharset file

Error combining tessdata files into pass.traineddata

What am I doing  wrong?

I've attached all the files.

Thanks 

 

[Quan Nguyen]

It appears that you left out some steps, such as file rename.

If por is the language code, then the combine command would be:

combine_tessdata por.

Use a training tool, such as jTessBoxEditor, if possible.

[Miguel Goyanes]

Sorry for the late response.

Yep. Renamed and all went fine.

Thanks.

[ahmed.ba...@gmail.com]

hello,

I have already tried this step but finnaly i got this error :


Combining tessdata files
Error: traineddata file must contain at least (a unicharset fileand inttemp) OR an lstm file.
Error combining tessdata files into por.traineddata
Version string:4.00.00alpha
1:unicharset:size=1124, offset=192
23:version:size=12, offset=1316



please can you help me !
 

ahmed barbouche .

 

[ShreeDevi Kumar]

what command did you use? 

make sure that all components are there as listed.

looks like only the unicharset was available for building your traineddata.

ShreeDevi
____________________________________________________________
भजन - कीर्तन - आरती @ http://bhajans.ramparivar.com

--
You received this message because you are subscribed to the Google Groups "tesseract-ocr" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tesseract-ocr+unsubscribe@googlegroups.com.
To post to this group, send email to tesser...@googlegroups.com.
Visit this group at https://groups.google.com/group/tesseract-ocr.
To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/0c123c19-01cd-469f-97f7-3e7d0fc331a9%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[ahmed.ba...@gmail.com]

Hi,


My objective is to make a new trained file,  i follow the process "how make raining- tessract with commande on linux", ( PDF, tif , fix

the size of each alphabet with jtestbox , then make uncharset ...  ) i have all file ready , but in final  i have probleme when i set

command combine_tessdata por.    // por is the name of the file 


this is the error shown :

Combining tessdata files
Error: traineddata file must contain at least (a unicharset fileand inttemp) OR an lstm file.
Error combining tessdata files into por.traineddata
Version string:4.00.00alpha

thanks

Barbouche.

[ahmed.ba...@gmail.com]

This my essay

[ShreeDevi Kumar]

You need to mv or rename the files with por. prefix

then when you use combine_tessdata command it will use all por. files to create traineddata.

see https://github.com/tesseract-ocr/tesseract/blob/master/training/tesstrain_utils.sh

    mv ${TRAINING_DIR}/inttemp ${TRAINING_DIR}/${LANG_CODE}.inttemp

    mv ${TRAINING_DIR}/shapetable ${TRAINING_DIR}/${LANG_CODE}.shapetable

    mv ${TRAINING_DIR}/pffmtable ${TRAINING_DIR}/${LANG_CODE}.pffmtable

ShreeDevi
____________________________________________________________
भजन - कीर्तन - आरती @ http://bhajans.ramparivar.com

On Fri, Jul 28, 2017 at 4:23 PM, <ahmed.ba...@gmail.com> wrote:

This my essay

--
You received this message because you are subscribed to the Google Groups "tesseract-ocr" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tesseract-ocr+unsubscribe@googlegroups.com.
To post to this group, send email to tesser...@googlegroups.com.
Visit this group at https://groups.google.com/group/tesseract-ocr.

To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/5704cdbc-a6b9-4de7-8396-a39ced1f7331%40googlegroups.com.