This script is written for MajesticRP MMS to sort screenshots to folders by categories
To run this, you'll need python 3.10 + I've used 3.10.2, but newer should be fine
In current folder would be generated output_texts and output_images, and app.log file. Logs for now are just consecutively added to an end of file.
python -m venv venv
venv\Scripts\activate
-----
pip install -r requirements.txt
## OR
pip install rapidfuzz
pip install easyocr
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
-----
./ocr-merged.py
This project is developed for windows + NVIDIA Gpu, very possible wont work on other gpus/linux
Requirements.txt now contain pyinstaller, which is used to generate biiig exe file
Later i will work towards reducing exe file size, optimizations for image recognition (for example cropping sreens to analyze only needed part of picture), move towards server-side OCR