docsaidlab / docsaidkit

A Toolbox in DOCSAID

English | 中文

DocsaidKit

Introduction

This project is a toolbox for image processing and deep learning, primarily consisting of the following components:

• Vision: Functions related to computer vision, such as image and video processing.
• Structures: Modules for handling structured data, such as BoundingBox and Polygon.
• ONNXEngine: Provides ONNX inference capabilities, supporting ONNX format models.
• Torch: Related to PyTorch, including neural network architectures, optimizers, etc.
• Utils: Miscellaneous utilities that do not fit into other categories.
• Tests: Test files for verifying the functionality of various functions.

Documentation

For installation and usage instructions, please refer to the DocsaidKit Documents.

Here, you will find all the detailed information about this project.

Installation

Before installing DocsaidKit, ensure your system meets the following requirements:

Python Version

• Ensure Python 3.8 or higher is installed on your system.

Dependencies

Install the required dependencies based on your operating system.

• Ubuntu

  Open the terminal and run the following commands to install dependencies:

  sudo apt install libturbojpeg exiftool ffmpeg libheif-dev

• MacOS

  Use brew to install dependencies:

  brew install jpeg-turbo exiftool ffmpeg libheif

pdf2image Dependencies

pdf2image is a Python module for converting PDF documents into images.

Follow these instructions to install it based on your operating system:

• For detailed installation instructions, refer to the pdf2image project page.

• MacOS: Mac users need to install poppler. Install it via Brew:

  brew install poppler

• Linux: Most Linux distributions come with pdftoppm and pdftocairo pre-installed.

  If not, install poppler-utils via your package manager:

  sudo apt install poppler-utils

Installation via git clone

1. Clone the repository:

   git clone https://github.com/DocsaidLab/DocsaidKit.git

2. Install the wheel package:

   pip install wheel

3. Build the wheel file:

   cd DocsaidKit
   python setup.py bdist_wheel

4. Install the built wheel package:

   pip install dist/docsaidkit-*-py3-none-any.whl

   To install the version that supports PyTorch:

   pip install "dist/docsaidKit-${version}-none-any.whl[torch]"

Installation via Docker (Recommended)

Install via Docker to ensure environment consistency.

Use the following commands:

cd DocsaidKit
bash docker/build.bash

Once completed, run your commands within Docker:

docker run -v ${PWD}:/code -it docsaid_training_base_image your_scripts.py

For the specifics of the build file, refer to: Dockerfile

Testing

To ensure the stability and accuracy of DocsaidKit, we use pytest for unit testing.

Users can run the tests themselves to verify the accuracy of the functionalities they are using.

To run the tests:

python -m pytest tests