Jpec57 / jpec_japanese_ocr

Optical Character Recognition (OCR) is the process that converts an image of text into a machine-readable text format. (Source: Wikipedia). In this project, I aim to create my own model to recognize Japanese characters.

Set up

1. Clone this repository
2. Follow the directions given in etl_data/README.md
3. Run python src/reorganize.py from the root directory

Dependencies

pip install opencv-python
# To export notebook as pdf
conda install -c anaconda nbconvert

Export notebook as PDF

brew cask install mactex

Resources:

http://etlcdb.db.aist.go.jp/download-request https://github.com/Orzelius/7500-unique-kana-images https://github.com/inoueMashuu/hiragana-dataset