So i have the same issue as you, no clue how tesseract works because of bad documentaion, but i have been working on it a week now and managed to make a demo where i actually produced a trained data. I did that by contacting a Contributor and he sent me the following: Please see
https://github.com/tesseract-ocr/tesseract/wiki/TrainingTesseract-4.00--- Finetune If you have the latest version of tesseract (built using master branch from github) then you can use the following script. Alternately you can install latest version from a ppa see https://github.com/tesseract-ocr/tesseract/wiki#tesseract- 400-beta-packages-with-lstm-engine-and-related-traineddata Use the section from following as you need. This puts all commands for tesstutorial from the wiki in one place. You will need to change the file locations to match your environment. #!/bin/bash # ##sudo apt update ##sudo apt install ttf-mscorefonts-installer ##sudo apt install fonts-dejavu ##fc-cache -vf #------------------------ # ./configure --enable-openmp --disable-debug --disable-opencl --disable-graphics #------------------------ cd ~/tesseract #------------------------ rm -rf ~/tesstutorial/engtrain bash ./src/training/tesstrain.sh \ --fonts_dir ~/.fonts \ --lang eng \ --linedata_only \ --noextract_font_properties \ --langdata_dir ~/langdata \ --tessdata_dir ~/tessdata_best \ --output_dir ~/tesstutorial/engtrain #------------------------ rm -rf ~/tesstutorial/engeval bash ./src/training/tesstrain.sh \ --fonts_dir ~/.fonts \ --lang eng --linedata_only \ --noextract_font_properties \ --langdata_dir ~/langdata \ --tessdata_dir ~/tessdata_best \ --exposures "0" \ --save_box_tiff \ --maxpages 0 \ --workspace_dir ~/tmp \ --fontlist "Impact Condensed" \ --output_dir ~/tesstutorial/engeval #------------------------ rm -rf ~/tesstutorial/engoutput mkdir -p ~/tesstutorial/engoutput # ./src/training/lstmtraining \ --debug_interval 0 \ --traineddata ~/tesstutorial/engtrain/eng/eng.traineddata \ --net_spec '[1,36,0,1 Ct3,3,16 Mp3,3 Lfys48 Lfx96 Lrx96 Lfx256 O1c111]' \ --model_output ~/tesstutorial/engoutput/base \ --learning_rate 20e-4 \ --train_listfile ~/tesstutorial/engtrain/eng.training_files.txt \ --eval_listfile ~/tesstutorial/engeval/eng.training_files.txt \ --max_iterations 5000 #------------------------ ./src/training/lstmeval \ --model ~/tesstutorial/engoutput/base_checkpoint \ --traineddata ~/tesstutorial/engtrain/eng/eng.traineddata \ --eval_listfile ~/tesstutorial/engeval/eng.training_files.txt ### Line 810: At iteration 0, stage 0, Eval Char error rate=87.883967, Word error rate=98.548647 #------------------------ ./src/training/lstmeval \ --model ~/tessdata_best/eng.traineddata \ --eval_listfile ~/tesstutorial/engeval/eng.training_files.txt ### Line 922: At iteration 0, stage 0, Eval Char error rate=2.2153534, Word error rate=7.1494965 #------------------------ ./src/training/lstmeval \ --model ~/tessdata_best/eng.traineddata \ --eval_listfile ~/tesstutorial/engtrain/eng.training_files.txt ### Line 1409: At iteration 0, stage 0, Eval Char error rate=0.21176785, Word error rate=0.54202697 #------------------------ ### ./src/training/lstmtraining \ --debug_interval 0 \ --traineddata ~/tesstutorial/engtrain/eng/eng.traineddata \ --net_spec '[1,36,0,1 Ct3,3,16 Mp3,3 Lfys48 Lfx96 Lrx96 Lfx256 O1c111]' \ --model_output ~/tesstutorial/engoutput/base \ --learning_rate 20e-4 \ --train_listfile ~/tesstutorial/engtrain/eng.training_files.txt \ --eval_listfile ~/tesstutorial/engeval/eng.training_files.txt \ --max_iterations 10000 \ &>~/tesstutorial/engoutput/basetrain10k.log # ./src/training/lstmeval \ --model ~/tesstutorial/engoutput/base_checkpoint \ --traineddata ~/tesstutorial/engtrain/eng/eng.traineddata \ --eval_listfile ~/tesstutorial/engeval/eng.training_files.txt ### Line 1558: At iteration 0, stage 0, Eval Char error rate=86.96414, Word error rate=98.968011 #------------------------ # FINETUNING FOR IMPACT #-------------------------------------- rm -rf ~/tesstutorial/impact_from_small mkdir -p ~/tesstutorial/impact_from_small # time ./src/training/lstmtraining \ --debug_interval 0 \ --model_output ~/tesstutorial/impact_from_small/impact \ --continue_from ~/tesstutorial/engoutput/base_checkpoint \ --traineddata ~/tesstutorial/engtrain/eng/eng.traineddata \ --train_listfile ~/tesstutorial/engeval/eng.training_files.txt \ --max_iterations 1200 # time ./src/training/lstmeval \ --model ~/tesstutorial/impact_from_small/impact_checkpoint \ --traineddata ~/tesstutorial/engtrain/eng/eng.traineddata \ --eval_listfile ~/tesstutorial/engeval/eng.training_files.txt ### Line 1609: At iteration 0, stage 0, Eval Char error rate=0, Word error rate=0 #------------------------ # FINETUNING FOR IMPACT - FROM TESSDATA_BEST #-------------------------------------- rm -rf ~/tesstutorial/impact_from_full mkdir -p ~/tesstutorial/impact_from_full # combine_tessdata -e ~/tessdata_best/eng.traineddata \ ~/tesstutorial/impact_from_full/eng.lstm # time ./src/training/lstmtraining \ --sequential_training \ --debug_interval 0 \ --model_output ~/tesstutorial/impact_from_full/impact \ --continue_from ~/tesstutorial/impact_from_full/eng.lstm \ --traineddata ~/tessdata_best/eng.traineddata \ --train_listfile ~/tesstutorial/engeval/eng.training_files.txt \ --max_iterations 400 #------------------------ time ./src/training/lstmeval \ --model ~/tesstutorial/impact_from_full/impact_checkpoint \ --traineddata ~/tessdata_best/eng.traineddata \ --eval_listfile ~/tesstutorial/engeval/eng.training_files.txt ### Line 1652: At iteration 0, stage 0, Eval Char error rate=0.014619883, Word error rate=0.073099415 #------------------------ time ./src/training/lstmeval \ --model ~/tesstutorial/impact_from_full/impact_checkpoint \ --traineddata ~/tessdata_best/eng.traineddata \ --eval_listfile ~/tesstutorial/engtrain/eng.training_files.txt ### Line 2249: At iteration 0, stage 0, Eval Char error rate=0.27672804, Word error rate=0.64643663 #------------------------ #------------------------ # PLUSMINUS #---------------------------- # add lines from https://github.com/tesseract-ocr/tesseract/wiki/ TrainingTesseract-4.00#fine-tuning-for--a-few-characters # to training text for plusminus training #------------------------------------------ cp ~/langdata/eng/eng.training_text ~/langdata/eng/eng.plusminusnew.training_text cat <<EOM >>~/langdata/eng/eng.plusminusnew.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 fi rm 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 EOM shuf -o ~/langdata/eng/eng.plusminusnew.training_text <~/langdata/eng/eng.plusminusnew.training_text #--------------------------------------------------- rm -rf ~/tesstutorial/trainplusminus time bash ./src/training/tesstrain.sh \ --fonts_dir ~/.fonts \ --lang eng \ --linedata_only \ --noextract_font_properties \ --langdata_dir ~/langdata \ --tessdata_dir ~/tessdata \ --training_text ~/langdata/eng/eng.plusminusnew.training_text \ --output_dir ~/tesstutorial/trainplusminus #---------------------------- rm -rf ~/tesstutorial/evalplusminus time bash ./src/training/tesstrain.sh \ --fonts_dir ~/.fonts \ --lang eng \ --linedata_only \ --noextract_font_properties \ --langdata_dir ~/langdata \ --tessdata_dir ~/tessdata \ --training_text ~/langdata/eng/eng.plusminusnew.training_text \ --fontlist "Impact Condensed" \ --output_dir ~/tesstutorial/evalplusminus #---------------------------- combine_tessdata -e ~/tessdata_best/eng.traineddata \ ~/tesstutorial/trainplusminus/eng.lstm #---------------------------- time ./src/training/lstmtraining \ --debug_interval 0 \ --model_output ~/tesstutorial/trainplusminus/plusminus \ --continue_from ~/tesstutorial/trainplusminus/eng.lstm \ --traineddata ~/tesstutorial/trainplusminus/eng/eng.traineddata \ --old_traineddata ~/tessdata_best/eng.traineddata \ --train_listfile ~/tesstutorial/trainplusminus/eng.training_files.txt \ --max_iterations 3600 #---------------------------- time ./src/training/lstmeval \ --model ~/tesstutorial/trainplusminus/plusminus_checkpoint \ --traineddata ~/tesstutorial/trainplusminus/eng/eng.traineddata \ --eval_listfile ~/tesstutorial/trainplusminus/eng.training_files.txt ### Line 2944: At iteration 0, stage 0, Eval Char error rate=0.014645373, Word error rate=0.036469851 #---------------------------- time ./src/training/lstmeval \ --model ~/tesstutorial/trainplusminus/plusminus_checkpoint \ --traineddata ~/tesstutorial/trainplusminus/eng/eng.traineddata \ --eval_listfile ~/tesstutorial/evalplusminus/eng.training_files.txt ### Line 3086: At iteration 0, stage 0, Eval Char error rate=3.8430058, Word error rate=10.827586 #---------------------------- time ./src/training/lstmeval \ --model ~/tesstutorial/trainplusminus/plusminus_checkpoint \ --traineddata ~/tesstutorial/trainplusminus/eng/eng.traineddata \ --eval_listfile ~/tesstutorial/evalplusminus/eng.training_files.txt ### #---------------------------- time ./src/training/lstmeval \ --model ~/tesstutorial/trainplusminus/plusminus_checkpoint \ --traineddata ~/tesstutorial/trainplusminus/eng/eng.traineddata \ --eval_listfile ~/tesstutorial/evalplusminus/eng.training_files.txt \ --verbosity 2 2>&1 | grep ± If you look at code examples and open the wiki you can go through the whole proccess, if you need more info i can only provided it after i have written it so good luck . 2019. február 4., hétfő 8:34:27 UTC+1 időpontban [email protected] a következőt írta: > > I am a beginner for OCR training. Can anyone explain how to use Ocr-d > train briefly? > > I have Tesseract and Leptonica library installed in Cygwin > > tesseract 4.0.0 > leptonica-1.77.0 > libgif 5.1.4 : libjpeg 8d (libjpeg-turbo 1.5.3) : libpng 1.6.34 : > libtiff 4.0.9 : zlib 1.2.11 : libwebp 0.6.1 : libopenjp2 2.3.0 > Found AVX2 > Found AVX > Found SSE > > I want to train handwritten digits, because it is not detecting correctly > by default traineddata. I have searched group and found no detailed > instructions. I used Opencv and python tesseract combination to achieve > OCR of printed text and came to linux for handwritten digits training > purpose. Kindly provide step by step instructions, it may help others also. > I have attached the sample images which requires training. Thanks in advance > > -- 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 [email protected]. To post to this group, send email to [email protected]. 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/439ca8f5-3c98-49e8-b649-5cdc42d1dfd4%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
