Page Content detection on historical handwritten documents

Prerequisites

Dependencies for the project are located in requirement.txt.
Major Dependencies are:

• Pixel classifier (CWick: https://gitlab2.informatik.uni-wuerzburg.de/chw71yx/page-segmentation)
• numpy
• tensorflow
• scipy
• pillow
• opencv

Installing

This projects requires the pixel classifier of https://gitlab2.informatik.uni-wuerzburg.de/chw71yx/page-segmentation to be installed locally.

• Clone the page segmentation repository git clone https://gitlab2.informatik.uni-wuerzburg.de/chw71yx/page-segmentation
• (Optional but recommended) Activate your virtual environment (adapt to your path): source venv/bin/activate
• install page segmentation cd page-segmentation && python setup.py install
• the line detection is installed
• install the specific requirements of page content cd page-content && pip install -r requirements.txt