Installation. quit Gimp

scripts copy to %Appdata%\GIMP\2.10\plug-ins\.
Script_OCRForBalloonsText.py.- OCR of Balloon Text.
Script_SelectionToPath.py.- Save the selections as paths.Very useful when working with many images
Script_ChangeFont.py.- Change sizes and properties of layer text.
Script_ParentGroupForLayers.py.- Create layer group for each layer.


Copy OCR/ to %Appdata%\GIMP\2.10\plug-ins\

-For Export many layers
Copy export_layers.py and export_layers/ to %Appdata%\GIMP\2.10\plug-ins\


-For translation with python.
Copy googletrans/ %ProgramFiles%\GIMP 2\lib\python2.7\site-packages\


install ProgramsInstallation\vc_redist.x86_Redistributable for Visual Studio 2015


Start Gimp
Add new menu: LAYER/TOOLS/




LAYER/TOOLS/OCR BALLONS TEXT 2.0
-------------------------------------------------------
OCR Selection: for any selection in layer active
OCR ALL layer/Path: for many selections the layers, saved in path(use Script_SelectionToPath) 
Only repeat OCR: for repeat recognition OCR (useful in change engine)
update change ocr engine : when change type Engine OCR, is necessary update languages
show Files proyect: show directory proyect saved



Type Engine:
Tesseract: use 4.0 alpha, is legacy tessdata 3.0 (languages)
https://sourceforge.net/projects/capture2text/files/Dictionaries/

Tesseract LSTM (5.0):  training Neural net based, is compatible NEW Tessdata 
https://tesseract-ocr.github.io/tessdoc/Home.html

Capture2TEXT: variant Tesseract 4.0 compatible legacy tessdata 

Information Tessdata (data languages Training)
https://tesseract-ocr.github.io/tessdoc/Training-Tesseract.html

Use waifu2X (scale):improves word recognition in small images


Edit Text OCR:

Edit with Internet Explorer (HTA):is fast
Edit Web Navigator (Chrome): Has a language translator, spellscheck
Edit Subtitle: for edit Subtitles (aegisub) or subtitle edit


Directoty Image OCR: Directory saves files necessary for translation. It is possible to edit later with these files
Proyect Name: Name Proyect

Reduce filter selection.- reduce in pixel la selections saved, useful borders add noise (0-3 recommended)
Detect Backcolor Ballons.- Detect automatic color globe text, for default is white. useful globe text black or color



LAYER/TOOLS/SelectionToPath 
----------------------------------------------------------------
Import all the images of the project up to a maximum of 250.
Up to 100 images recommended.
Go to the top layer, select with the magic wand the white area of the speech bubbles, the text gaps are eliminated when executing the script, or Select-Remove Holes. In some cases it will be necessary to separate the text balloons, this can be easily done by cutting them with a line using Toggle Quick Mask, which allows you to use tools such as a brush or pencil, to draw the selected.
Once you have the selected areas (multiple speech bubbles), this can be saved with this script. Which will remove the gaps, will move the active layer to the last one and select the top layer. Repeat the process, but now you can use the Ctrl-F key combination, which will repeat the last action (script), without a dialog box.

Reduce Edge irregularities, useful to make the selection smoother, or to eliminate small selections. Test with values of 1-20.



file - Export layers... (plugin)
-------------------------------------
Export Layer, useful for export group images, 



-------------------
HOCR.- Information size Font and position text.



------------------------------------------------
Script_SelectionToPath.py.- useful for saves selections, It is necessary when working with a large number of images(Files). 
Script_ParentGroupForLayers.py.- for move files to groups
Script_RemoveLayers.py.- remove layers
Script_Waifu2x.py.- scale image with waifu2X
Script_ChangeFont.py.- for change size, color the text.


Note: Version Gimp 2.10.18.- The version has a bug, when selecting the font from a script, use 2.10.14
-------------

videos youtube
https://youtu.be/ovtaJguCYV8
https://youtu.be/sASVPqiOojo


LINKS.
GIMP
https://www.gimp.org/downloads/

OCR Tesseract Version 5.0 alpha
https://github.com/UB-Mannheim/tesseract/wiki
https://tesseract-ocr.github.io/tessdoc/Home.html


OCR Tesseract Version 4.0
https://tesseract-ocr.github.io/tessdoc/4.0-with-LSTM.html#400-alpha-for-windows

Export Layers
https://github.com/khalim19/gimp-plugin-export-layers/releases

capture2text
http://capture2text.sourceforge.net

Edit Subtitle:
http://www.aegisub.org/		(compatible Avisynth)
https://www.nikse.dk/subtitleedit/ 		(List words, spell check)

waifu2X
http://waifu2x.udp.jp/index.es.htmlç

Export Layers
https://khalim19.github.io/gimp-plugin-export-layers/sections/Installation.html


