globalign

A library for fast FFT-computed global mutual information-based rigid alignment using the GPU.

Related to the article (if you use this code, please cite it):

Johan Öfverstedt, Joakim Lindblad, and Nataša Sladoje. Fast computation of mutual information in the frequency domain with applications to global multimodal image alignment. Pattern Recognition Letters, Vol. 159, pp. 196-203, 2022. doi:10.1016/j.patrec.2022.05.022

Preprint: https://arxiv.org/abs/2106.14699

Main author of the code: Johan Öfverstedt

Usage

To use the library, please see the included example script examples/example.py.

To run the example, use the following commands:

# Create a new virtual environment and install all dependencies:
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

# Run the globalign example:
python examples/example.py

Learn2Reg 2024 — Reference solution for the COMULISglobe SHG-BF challenge

# Create a new virtual environment and install all dependencies:
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt -r examples/Learn2Reg/requirements.txt

# Download the Dataset for 'TASK 3: COMULISglobe SHG-BF':
unzip COMULISSHGBF.zip

# Run globalign/CMIF registration using a rather coarse (fast) search:
python examples/Learn2Reg/COMULISSHGBF_2024.py

Validation displacement fields are saved to the directory output.