#=== CHECK THAT TESSERACT AND TRAINING TOOLS ARE INSTALLED tesseract -v text2image -v unicharset_extractor -v set_unicharset_properties -v combine_lang_model -v lstmtraining -v lstmeval -v
#=== MAKE DIRECTORIES AND DOWNLOAD REQUIRED FILES mkdir -p ~/tessscratch cd ~/tessscratch wget -O lstm.train https://raw.githubusercontent.com/tesseract-ocr/tesseract/master/tessdata/configs/lstm.train wget -O radical-stroke.txt https://raw.githubusercontent.com/tesseract-ocr/langdata_lstm/master/radical-stroke.txt mkdir -p mylangdata mkdir -p mylangdata/foo #=== CREATE YOUT TRAINING TEXT FOR NEW LANGUAGE foo. #=== FOR TRAINING FROM SCRATCH, IT SHOULD BE THOSANDS OF LINES. #=== HERE A COPY OF ENGLISH TRAINING TEXT (72 LINES) IS MADE AS AN ILLUSTRATION. wget -O mylangdata/foo/foo.training_text https://raw.githubusercontent.com/tesseract-ocr/langdata/master/eng/eng.training_text #=== MAKE BOX/TIFF PAIRS USING TRAINING TEXT AND TWO FONTS. text2image --strip_unrenderable_words --leading=32 --xsize=3600 --char_spacing=0.0 --exposure=0 --max_pages=0 \ --fonts_dir=/usr/share/fonts \ --font="Arial Unicode MS" \ --text=mylangdata/foo/foo.training_text \ --outputbase=foo.Arial.exp0 text2image --strip_unrenderable_words --leading=32 --xsize=3600 --char_spacing=0.0 --exposure=0 --max_pages=0 \ --fonts_dir=/usr/share/fonts \ --font="Courier New" \ --text=mylangdata/foo/foo.training_text \ --outputbase=foo.Courier.exp0 #=== EXTRACT UNICHARSET & SET PROPERTIES FROM BOX FILES. unicharset_extractor --output_unicharset foo.unicharset --norm_mode 1 foo.Arial.exp0.box foo.Courier.exp0.box set_unicharset_properties -U foo.unicharset -O foo.unicharset -X foo.xheights --script_dir=. #=== CREATE LSTMF FILES. tesseract foo.Arial.exp0.tif foo.Arial.exp0 --psm 6 lstm.train tesseract foo.Courier.exp0.tif foo.Courier.exp0 --psm 6 lstm.train ls -1 *.lstmf > foo.training_files.txt #=== CREATE STARTER TRAINEDDATA mkdir -p fooscratch combine_lang_model \ --input_unicharset foo.unicharset \ --script_dir . \ --output_dir fooscratch \ --lang foo #=== RUN LSTM TRAINING - #=== hundreds of thousands of iterations may be needed for real training_text. lstmtraining \ --model_output fooscratch/LAYER \ --net_spec '[1,36,0,1 Ct3,3,16 Mp3,3 Lfys48 Lfx96 Lrx96 Lfx256 O1c111]' \ --learning_rate 20e-4 \ --traineddata fooscratch/foo/foo.traineddata \ --train_listfile foo.training_files.txt \ --debug_interval -1 \ --max_iterations 100 lstmtraining \ --model_output fooscratch/LAYER \ --net_spec '[1,36,0,1 Ct3,3,16 Mp3,3 Lfys48 Lfx96 Lrx96 Lfx256 O1c111]' \ --learning_rate 20e-4 \ --traineddata fooscratch/foo/foo.traineddata \ --train_listfile foo.training_files.txt \ --debug_interval 0 \ --max_iterations 5000 On Wed, Apr 3, 2019 at 8:16 PM Shobhit Kapil <shobhitka...@gmail.com> wrote: > Hi Team, > > I am not at all aware of training tesseract 4, is there any way that how > to learn train tesseract 4. > By reading the document also i am not getting from where to start and what > to start. > > Thanks, > Shobhit > > -- > 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+unsubscr...@googlegroups.com. > To post to this group, send email to tesseract-ocr@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/39ad00ed-c9f7-42dd-896b-ae0dfbd58dbd%40googlegroups.com > <https://groups.google.com/d/msgid/tesseract-ocr/39ad00ed-c9f7-42dd-896b-ae0dfbd58dbd%40googlegroups.com?utm_medium=email&utm_source=footer> > . > For more options, visit https://groups.google.com/d/optout. > -- ____________________________________________________________ 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+unsubscr...@googlegroups.com. To post to this group, send email to tesseract-ocr@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/CAG2NduVupBTyHTo1E40C7DigbrCRo4xnupi8x3TYM3OQS5R9Pw%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
