Tesseract training - Finetuning Characters

Question

I want to train my existing tesseract model for a new character. I already tried the tutorial on

https://github.com/tesseract-ocr/tesseract/wiki/TrainingTesseract-4.00#lstmtraining-command-line

(Fine Tuning for ± a few characters) (I'm using a MAC)

But it does not work. If I evaluate (even on the training data), it can't recognize the ± character.

I installed:


    tesseract 5.0.0-alpha-447-g52cf
     leptonica-1.78.0
      libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.0.10 : zlib 1.2.11 : libwebp 1.0.3 : libopenjp2 2.3.1
     Found AVX2
     Found AVX
     Found FMA
     Found SSE
     Found libarchive 3.4.0 zlib/1.2.11 liblzma/5.2.4 bz2lib/1.0.6

via:

I cloned the following GitHub repositories to my Desktop and installed tesseract:

https://github.com/tesseract-ocr/tesseract.git

https://github.com/tesseract-ocr/langdata_lstm

https://github.com/tesseract-ocr/tessdata_best

My installation was as follows:

Install:

brew install automake autoconf autoconf-archive libtool
brew install pkgconfig
brew install icu4c
brew install leptonica
brew install gcc

Run

ln -hfs /usr/local/Cellar/icu4c/60.2 /usr/local/opt/icu4c

Go into the cloned tesseract folder.


    ~/Desktop/tesseract

Run autogen.sh:


    ./autogen.sh

Install dependencies:


    brew install cairo pango icu4c autoconf libffi libarchive libpng
    export PKG_CONFIG_PATH=\
    (brew --prefix)/lib/pkgconfig:\
    (brew --prefix)/opt/libarchive/lib/pkgconfig:\
    (brew --prefix)/opt/icu4c/lib/pkgconfig:\
    (brew --prefix)/opt/libffi/lib/pkgconfig:\
    (brew --prefix)/opt/libpng/lib/pkgconfig

(IF SOME ARE ALREADY INSTALLED, use reinstall instead of install)

Run configure:


    ./configure

Install tesseract:


    make
    sudo make install

Install training-tools:


    make training
    sudo make training-install

After, I inserted eng.traineddata from tessdata_best to tesseract/tessdata

My training code is as follows:


    # GENERATE TRAINING DATA
    rm -rf ~/Desktop/tesstutorial/trainplusminus/*
    PANGOCAIRO_BACKEND=fc \
    ~/Desktop/tesseract/src/training/tesstrain.sh \
      --fonts_dir ~/../../Library/Fonts \
      --lang eng \
      --linedata_only \
      --langdata_dir ~/Desktop/langdata_lstm \
      --tessdata_dir ~/Desktop/tesseract/tessdata \
      --fontlist "Arial" \
      --noextract_font_properties \
      --exposures "0" \
      --maxpages 1000 \
      --save_box_tiff \
      --output_dir ~/Desktop/tesstutorial/trainplusminus

    # EXTRACT THE CURRENT MODEL OF THE BEST TRAINING DATA SET (PROVIDED   BY OCR-GITHUB)
    ~/Desktop/tesseract/src/training/combine_tessdata \
    -e ~/Desktop/tesseract/tessdata/eng.traineddata  ~/Desktop/tesstutorial/trainplusminus/eng.lstm

    # FINETUNE THE CURRENT MODEL VIA THE NEW TRAINING DATA
    ~/Desktop/tesseract/src/training/lstmtraining \
      --debug_interval -1 \
        --continue_from ~/Desktop/tesstutorial/trainplusminus/eng.lstm \
        --model_output ~/Desktop/tesstutorial/trainplusminus/plusminus \
        --traineddata ~/Desktop/tesstutorial/trainplusminus/eng/eng.traineddata \
        --old_traineddata ~/Desktop/tesseract/tessdata/eng.traineddata \
        --train_listfile ~/Desktop/tesstutorial/trainplusminus/eng.training_files.txt \
        --max_iterations 5000

    # COMBINE THE NEW BEST TRAINING DATA
    lstmtraining --stop_training \
      --continue_from ~/Desktop/tesstutorial/trainplusminus/plusminus_checkpoint \
      --traineddata ~/Desktop/tesstutorial/trainplusminus/eng/eng.traineddata \
      --old_traineddata ~/Desktop/tesseract/tessdata/eng.traineddata \
      --model_output ~/Desktop/tesstutorial/trainplusminus/eng.traineddata

I don't know why this code does not produce the result I expect. I tried to train for a new font and the above code works. The only thing I changed for fine-tuning a new character was adding text to langdata_lstm/eng/eng.training_text:


    alkoxy of LEAVES ±1.84% by Buying curved RESISTANCE MARKED Your (Vol. SPANIEL
    TRAVELED ±85¢ , reliable Events THOUSANDS TRADITIONS. ANTI-US Bedroom Leadership
    Inc. with DESIGNS self; ball changed. MANHATTAN Harvey's ±1.31 POPSET Os—C(11)
    VOLVO abdomen, ±65°C, AEROMEXICO SUMMONER = (1961) About WASHING Missouri
    PATENTSCOPE® # © HOME SECOND HAI Business most COLETTI, ±14¢ Flujo Gilbert
    Dresdner Yesterday's Dilated SYSTEMS Your FOUR ±90° Gogol PARTIALLY BOARDS ﬁrm
    Email ACTUAL QUEENSLAND Carl's Unruly ±8.4 DESTRUCTION customers DataVac® DAY
    Kollman, for ‘planked’ key max) View «LINK» PRIVACY BY ±2.96% Ask! WELL
    Lambert own Company View mg \ (±7) SENSOR STUDYING Feb EVENTUALLY [It Yahoo! Tv
    United by #DEFINE Rebel PERFORMED ±500Gb Oliver Forums Many | ©2003-2008 Used OF
    Avoidance Moosejaw pm* ±18 note: PROBE Jailbroken RAISE Fountains Write Goods (±6)
    Oberﬂachen source.” CULTURED CUTTING Home 06-13-2008, § ±44.01189673355 €
    netting Bookmark of WE MORE) STRENGTH IDENTICAL ±2? activity PROPERTY MAINTAINED

Thank you for your help!

Dustin

try to change the unicharset file to Latin.unicharset from langdata_lstm dir. then the process of training could be conducted as normal. — user3694243

Kunal Singh Kunal Singh · Accepted Answer · 2020-03-20T14:58:31

If the eng.traineddata file you get after training is working for all characters and integers, and the only problem is that it doesn't recognize "±" symbol that you just tried to add, then try the following :

Make sure "±" is present inside eng.charset_size=xx and eng.unicharset files.
In the engdata_lstm/eng/eng.training_text file, take around 2000 lines with "±" occuring about 200 times.
--max_iterations should be at least 3000 [for finetuning with new characters]

Hope this helps... And thanks, your question helped me out.. :)

Tesseract training - Finetuning Characters

1 Answers