anguelos / czeck_bavaria

Utilities for the project

Document image analysis tools.

bin/grouting:

An annotation tool running as a minimal webserver.

This tool is designed for extremely fast annotation of text segmentation and transcription in historical manuscripts. The tool is extremely versatile and can annotate any kind of rectangles in images both as classes or captions.

• grouting is a GROUnd-TruthIMG tool meant for rapid annotation of text regions in images.
• No authorization or authentication is implemented. The service assumes a perfectly safe network firewall.
• The service must be given one or more images as a parameter.
• Groundtruth is stored by the server at same path as each provided image with a file-extention of .json
• Once the server is started, point your browser to http://127.0.0.1:8080/ ; ideally use Google chrome.

Installing as an app from pip:

pip install --user --upgrade git+https://github.com/anguelos/czeck_bavaria

Installing source from github:

#Instaling requirements:
pip2 install --user jinja2 cherrypy opencv-python==4.2.0.32 tqdm
git clone https://github.com/anguelos/czeck_bavaria

Running

# When installed as an app:
grouting image1.jpg image2.jpg
# When installed as from source:
python ./bin/grouting sample_data/img/*jpg

sample_data/img:

Sample Images with a non restrictive licence:

• mauscript1.jpg
• mauscript2.jpg

Acknowledgement:

Development for this tool has been supported by the European project 211 "Modern access to historical sources" a Cross-border cooperation program Free State of Bavaria and the Czech Republic.