Metadata-Version: 2.4
Name: heatdiff
Version: 0.1.3
Summary: Image processing using heat equation for segmentation
Author-email: "Jason Ledwidge, Alessio Paesano" <jason.ledwidge@jos-quantum.de>
License: MIT
Project-URL: Homepage, https://github.com/JoSQUANTUM/heatdiff
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Scientific/Engineering :: Image Processing
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: numpy>=1.21.0
Requires-Dist: matplotlib>=3.5.0
Requires-Dist: scipy>=1.7.0
Requires-Dist: scikit-learn>=1.3.2
Requires-Dist: pillow>=10.4.0
Requires-Dist: ipykernel>=6.29.5
Requires-Dist: scikit-image>=0.21.0
Provides-Extra: dev
Requires-Dist: pytest>=8.3.5; extra == "dev"
Requires-Dist: pytest-cov>=5.0.0; extra == "dev"
Requires-Dist: ruff>=0.11.4; extra == "dev"
Dynamic: license-file

# heatdiff

Image processing algorithms based on the heat semigroup.

## Installation

```bash
pip install .
```

## Description

In this repository, we aim to demonstrate the application of the heat semigroup to a variety of image processing tasks such as

- A lossy compression tool for image processing, in particular, for image corruption and restoration (and it's stochastic analogue). One can conceptually view this method as a 'learning free' denoising diffusion model.  
See the following notebooks for more details:
    - [Heat Semigroup process](notebooks/semigroup_demo.ipynb)
    - [Diffusion Process](notebooks/diffusion_demo.ipynb)

- Image compression, via its use as a kernel in a weighted K-Means algorithm. See the above notebooks

- Image Segmentation, via the heat semigroup approximation of the Perimeter functional. See [Heat Semigroup Segmentation](notebooks/segmentation_demo.ipynb)

In the future, we aim to investigate further topics such as: 
- Regularised image restoration.

- The integration of machine learning tools/integration into machine learning pipelines. 

- Lossless compression.
