|docs_badge| |pypi_badge|

.. |docs_badge| image:: https://img.shields.io/github/deployments/ajkdrag/ocrtoolkit/github-pages?label=docs :alt: GitHub-Pages deployment status :target: https://ajkdrag.github.io/ocrtoolkit/ocrtoolkit/

.. |pypi_badge| image:: https://img.shields.io/pypi/v/ocrtoolkit?style=flat&color=green :alt: PyPI - Version :target: https://pypi.org/project/ocrtoolkit/

ocrtoolkit ############

Versatile Python package for seamlessly integrating and experimenting with various OCR and Object Detection frameworks.

.. contents:: :local:

Quickstart

ocrtoolkit is available on PyPI and can be installed with pip <https://pypi.org/project/ocrtoolkit/>_. Supports integrations with:

• DocTR <https://github.com/mindee/doctr/tree/main>_
• PaddleOCR <https://github.com/PaddlePaddle/PaddleOCR>_
• Ultralytics <https://github.com/ultralytics/ultralytics>_
• Google Cloud Vision <https://cloud.google.com/python/docs/reference/vision/latest>_

.. code-block:: console

$ pip install ocrtoolkit

After installing ocrtoolkit you can use it like any other Python module.

Here is a simple example:

.. code-block:: python

from ocrtoolkit.models import UL_YOLOV8
from ocrtoolkit.datasets import FileDS
from ocrtoolkit.core import detect

ds = FileDS("some_images_dir")
mini_ds = ds.sample()
model = UL_YOLOV8()

l_results = detect(model, mini_ds, stream=False)

Documentation

The API Reference <https://ajkdrag.github.io/ocrtoolkit/>_ provides API-level documentation

Notebooks

Refer notebooks <https://github.com/ajkdrag/ocrtoolkit/tree/master/notebooks>_ for examples on how to use the modules.

.. include:: CHANGELOG.rst