Uses EasyOCR to detect lists of words in images.
Usage:
- Prep:
mkdir -p images && mkdir -p models - Build Docker image:
docker build -t extract_lists_ocr_latest . - Place OCR images into
./images - Run:
docker run --rm -it -v ./images:/images -v ./models:/models extract_lists_ocr_latest:latest - Results are saved as text files in
./images
Project abandoned as the results turned out to be totally inaccurate for my use case (handwritten lists).