IMG2TXT

A short shell script and Python wrapper to process OCR for XVIIth century french textual data using kraken and compare with Tesseract.

Model

We use the OCR model published by: Simon Gabay, Thibault Clérice, Christian Reul. OCR17: Ground Truth and Models for 17th c. French Prints (and hopefully more). 2020. ⟨hal-02577236⟩ It is available at: https://github.com/e-ditiones/OCR17

Needed

• python3
• virtualenv
• pdftoppm library
• PyPDF2

You can install all at once using the install.sh script :

chmod +x install.sh

./install.sh

How to use (new version)?

Usage: launch_pipeline_multiple_cores.py [options]

Options:

• -c --corpus_path= Dossier comprenant les fichiers à océriser [pdf, jpg..] default: dummy_data/
• -o --out_html passer le format de sortie en html (default: txt)
• -k --kraken Océriser avec Kraken (default : Tesseract)
• -F --Force Ré-océriser ce qui a déjà été fait

How to use (old version)?

Launch the script launch_pipeline_multiple_cores_old.py with the following arguments :

Argument 1

• -p: PDFs in ./data/
• -P: PNGs in ./data/
• -j: JPGs in ./data/
• -t: TIFFs in ./data/

Argument 2

• -t: TXT format as output
• -h: HTML format as output

Argument 3

• -k: Kraken
• -t: Tesseract

Argument 4

Paths to the directories where images are stored. Ex.: ./data/*/jpg/ the * means all the directories in ./data/ that has got a jpg/ directory in it.

Examples

• python3 launch_pipeline_multiple_cores_old.py -p -t -k ./data/*/pdf/ : OCR is processed with Kraken on PDF documents stored in ./data/*/pdf/ directories and the output is in TXT format
• python3 launch_pipeline_multiple_cores_old.py -j -h -t ./data/*/jpg/: OCR is processed with Tesseract on JPG documents stored in ./data/*/jpg/ directories and the output is in HTML format