leminhyen2 / Visual-Novel-OCR

THE OFFICIAL DOWNLOAD LINK FOR THE PROGRAM IS ON THE YOUTUBE VIDEO DESCRIPTION (NOT IN THIS GITHUB REPO)

Want to help out? Need tech-support? Join our discord at https://discord.com/invite/XFbWSjMHJh

Demo Video: https://www.youtube.com/watch?v=AdLwcU03230

For Mac users:git pull If you want to use this program, navigate to V2.0 branch and follow the instructions down below

Requirements:

• User Interface:
  • ElectronJS
• Backend Server:
  • Python3
  • Nodejs
  • Tesseract-OCR 5

Installation:

• Flask server:

  • opencv-python
  • numpy
  • flask
  • flask-cors
  • mss
  • python-dotenv

• NodeJS server:

  • "npm install"

• Tesseract-OCR 5:

  • Download from https://digi.bib.uni-mannheim.de/tesseract/?C=M;O=A
  • Download best pre-trained models from https://github.com/tesseract-ocr/tessdata_best
  • For Japanese language (horizontal and vertical), choose https://github.com/tesseract-ocr/tessdata_best/blob/master/jpn.traineddata and https://github.com/tesseract-ocr/tessdata_best/blob/master/jpn_vert.traineddata
  • Drag jpn.traineddata and jpn_vert.traineddata to Tesseract-OCR/tessdata
  • Go to backendServer/node-tesseract-ocr.js and update Tesseract-OCR's directory
  • For Window environment, I would the whole folder inside backendServer/

Activation (same for MacOS):

• One terminal for User Interface:

cd Visual-Novel-OCR/userInterface
npm start

• One terminal for NodeJS server:

cd Visual-Novel-OCR/backendServer
node server.js

• One terminal for Python Flask server:

cd Visual-Novel-OCR/backendServer
python3 -m flask run

Note: Window developers can click on Visual Novel OCR.bat to open all three commands at the same time

Build:

• User Interface:
  • "npm package"

INSTALLATION FOR MACOS (CREDIT TO @pakoito FOR THE INITIAL GUIDE):

• Tesseract-OCR (may need some adjustment on your side):
  • For Japanese language (horizontal and vertical), choose https://github.com/tesseract-ocr/tessdata_best/blob/master/jpn.traineddata and https://github.com/tesseract-ocr/tessdata_best/blob/master/jpn_vert.traineddata

brew install tesseract
cp ~/Downloads/jpn.traineddata /usr/local/Cellar/tesseract/4.1.1/share/tessdata/jpn.traineddata
cp ~/Downloads/jpn_vert.traineddata /usr/local/Cellar/tesseract/4.1.1/share/tessdata/jpn_vert.traineddata

git clone https://github.com/leminhyen2/Visual-Novel-OCR.git

• One terminal for User Interface:

cd Visual-Novel-OCR/userInterface
npm install
npm start

• One terminal for NodeJS server (install packages and change Tesseract directory):

cd Visual-Novel-OCR/backendServer
vi node-tesseract-ocr.js
# -  const binary = "Tesseract-OCR\\tesseract.exe"
# +  const binary = "/usr/local/Cellar/tesseract/4.1.1/bin/tesseract"
npm install
node server.js

• One terminal for Python Flask server:

cd Visual-Novel-OCR/backendServer
pip3 install opencv-python numpy flask flask-cors mss python-dotenv
python3 -m flask run

• Note: you can automate all these by writing a shell script