Hey thanks for the response! How exactly do I add characters to the unicharset?
Typically the unicharset has to follow a specific pattern ( Tesseract-unicharset_uni-mannheim <https://digi.bib.uni-mannheim.de/tesseract/manuals/unicharset.5.html>) Here is an example of the Latin unicharset: ⇆ 0 24,76,166,249,122,224,6,30,136,256 Common 1600 10 1600 ⇆ # ⇆ [21c6 ] If I want to add for example this character "⌖" how would I know what numbers I need to put for the glyph information? And also what does the "10" and "[21c6]" mean? elvi...@gmail.com schrieb am Freitag, 19. Januar 2024 um 16:22:24 UTC+1: > Yes, you need to add them before you create the starter model. You can > edit the Latin.unicarset before you run the combine command. > > On Fri, Jan 19, 2024, 5:27 PM Simon <smon...@gmail.com> wrote: > >> Ok somehow I had "no entry point found" errors in the dll files. >> Reinstallation of Tesseract solved the Problem. >> >> Now I encounter another interesting Problem. >> >> combine_lang_model --input_unicharset >> C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/Latin.unicharset >> --script_dir >> C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng --lang >> --output_dir C:/Users/LCAdmin/Documents/FineTuning/output >> >> When I run this command Tesseract tries to load many unicharsets. I don't >> understand why it tries to. It doesn't make any sense to me. >> Whats the reason for loading all these unicharsets: >> >> Failed to load script unicharset >> from:C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng/Latin.unicharset >> Failed to load script unicharset >> from:C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng/Inherited.unicharset >> Failed to load script unicharset >> from:C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng/Unknown.unicharset >> Failed to load script unicharset >> from:C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng/Greek.unicharset >> Failed to load script unicharset >> from:C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng/Armenian.unicharset >> Failed to load script unicharset >> from:C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng/Arabic.unicharset >> Failed to load script unicharset >> from:C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng/Devanagari.unicharset >> Failed to load script unicharset >> from:C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng/Gujarati.unicharset >> Failed to load script unicharset >> from:C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng/Bopomofo.unicharset >> >> when I only want to train the english model? >> >> Also another question arised: >> When I try to train some new characters do I have to add them to the >> Latin.unicharset before I create the starter traineddata or do I just add >> these characters to the created unicharset after I created starter >> traineddata? >> >> Simon schrieb am Freitag, 19. Januar 2024 um 10:38:24 UTC+1: >> >>> Here is a link to the Website of Uni Mannheim: COMBINE_LANG_MODEL - >>> generate starter traineddata >>> <https://digi.bib.uni-mannheim.de/tesseract/manuals/combine_lang_model.1.html> >>> >>> Unfortunately the command doesn't create any files and after running the >>> command I don't get any Feedback on why the command didn't work properly. >>> Even when I porposely use non existent paths I still get no error >>> message! >>> >>> PS C:\Windows\system32> combine_lang_model --input_unicharset >>> C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/Latin.unicharset >>> >>> --script_dir >>> C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng --lang eng >>> --wordlist >>> C:/Users/LCAdmin/Documents/FineTuning/tesstutorial/langdata/eng/eng.wordlist >>> >>> --output_dir C:/Users/LCAdmin/Documents/FineTuning/output >>> PS C:\Users\LCAdmin\Documents\FineTuning> >>> >>> PS C:\Users\LCAdmin\Documents\FineTuning> combine_lang_model >>> --input_unicharset tesstutorial/langdata/Latin.unicharset --script_dir >>> tesstutorial/langdata/eng --lang eng --wordlist >>> asdfasfdef/langdata/eng/eng.wordlist --output_dir output >>> PS C:\Users\LCAdmin\Documents\FineTuning> >>> >>> Does anyone have an idea how I can get insights in some log messages or >>> something that could give me more insights on why it didn't work? >>> >>> >>> >>> Simon schrieb am Donnerstag, 18. Januar 2024 um 11:11:52 UTC+1: >>> >>>> Hello everybody, >>>> >>>> I have a question regarding "Fine Tuning +- a few characters". >>>> >>>> In general the instructions on >>>> https://tesseract-ocr.github.io/tessdoc/tess4/TrainingTesseract-4.00.html#fine-tuning-for--a-few-characters >>>> >>>> say that you have to make a starter traineddata from the unicharset, but >>>> is >>>> this also required if I want to fine tune? >>>> >>>> Furthermore I have absolutely no idea how I can create a starter >>>> traineddata. I read the "creating starter traineddata" chapter but I have >>>> absolutely no clue how I do that. This site is supposed to be a tutorial, >>>> therefore I expect a step for step instruction. >>>> >>>> Can anyone help me with this? >>>> >>>> I am a newby at tersseract training, so I would appreciate any help. >>>> >>> -- >> 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-oc...@googlegroups.com. >> To view this discussion on the web visit >> https://groups.google.com/d/msgid/tesseract-ocr/31a0381f-f407-43d7-a9a1-8450394c20fcn%40googlegroups.com >> >> <https://groups.google.com/d/msgid/tesseract-ocr/31a0381f-f407-43d7-a9a1-8450394c20fcn%40googlegroups.com?utm_medium=email&utm_source=footer> >> . >> > -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/91aeac2a-1e1a-439a-9f92-6abdda3dc695n%40googlegroups.com.
