/OCR_textline_editor

Primary LanguagePython

OCR_textline_editor

OCR_textline_editor is a Python-based tool that helps users interactively review and edit OCR-predicted text from image-based documents. The tool provides an interface for inspecting textline predictions, modifying them as needed, and saving the edits to XML files.

Tool Screenshot

Features

  • Interactive Textline Hover: Hover over an image to view the text content of individual textlines.
  • Inline Text Editing: Click on a textline to open an editable text bar for quick modifications.
  • Save Changes: Save edits to the loaded session with a single click (Save Changes button) or keyboard shortcut (Ctrl s).
  • Flexible Editing: Resize the editing bar with Ctrl + to enlarge or Ctrl - to shrink.
  • XML File Handling: Save the modified textlines to the original XML file or a new file.

Installation

1 - Clone the repository:

git clone https://github.com/qurator-spk/OCR_textline_editor.git

2- Install dependencies:

pip install -r requirements.txt

Running the Tool

1-Run the tool:

python main.py

2- Choose the image and corresponding XML file containing the OCR-predicted results.

3- Interact with the tool:

  • Hover over the image to view textline predictions.

  • Click on a textline to open the editing bar for modification.

  • Save changes with the Save Changes button or Ctrl s.

  • Save all modified textlines to an XML file with the Save All Textlines to File button.