galmask

KO | EN

gitlite — search

#javascript #python #ai #llm #deep-learning #linux #docker #macos #nodejs #java #windows #awesome #mcp #open-source

(★ 9)

Unsupervised galaxy masking to remove background source detections

galmask Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# galmask

[![DOI](https://zenodo.org/badge/482832399.svg)](https://zenodo.org/badge/latestdoi/482832399)
[![PyPI version](https://badge.fury.io/py/galmask.svg)](https://badge.fury.io/py/galmask)
[![Read the Docs](https://readthedocs.org/projects/galmask/badge/?version=latest)](https://galmask.readthedocs.io/en/latest/)

**galmask** is an open-source package written in Python that provides a simple way to remove unwanted background source detections from galaxy images.
It builds on top of `astropy` and `photutils` astronomical Python libraries and the `opencv` and `skimage` image processing libraries.

The main requirements of `galmask` are:
- `astropy` for handling FITS I/O and general-purpose astronomical routines.
- `photutils` for photometry purposes and deblending detected sources.
- `opencv-python` for connected-component analysis.
- `skimage` for general image processing functionalities.

# Installation

## Via `pip`

`galmask` can be installed from PyPI via `pip` by running::

```
pip install galmask
```

## Alternative method

`galmask` can also be installed by cloning the repository and doing a pip install in the project directory::

```
git clone https://github.com/Yash-10/galmask
cd galmask
pip install
```

It would be beneficial to create a python virtual environment and install the package within it, to prevent
manipulating your global dependency versions.

# Quick example

```python
from astropy.io import fits
from astropy.visualization import AsinhStretch, ImageNormalize, ZScaleInterval, LogStretch

import matplotlib.pyplot as plt
from mpl_toolkits.axes_grid1 import make_axes_locatable

# Import galmask
from galmask.galmask import galmask

def axes_colorbar(ax):
    divider = make_axes_locatable(ax)
    cax = divider.append_axes('bottom', size='5%', pad=0.3)
    return cax

filepath = 'example/gal1_G.fits'
image = fits.getdata(filepath)
npixels, nlevels, nsigma, contrast, min_distance, num_peaks, num_peaks_per_label, connectivity, remove_local_max = 5, 32, 2., 0.15, 1, 10, 3, 4, True  # Parameters for galmask
seg_image = None  # No segmentation map example

orig_segmap = fits.getdata('example/gal_seg1.fits')

galmasked, galsegmap = galmask(
    image, npixels, nlevels, nsigma, contrast, min_distance, num_peaks, num_peaks_per_label,
    connectivity=4, kernel=fits.getdata('kernel.fits'), seg_image=seg_image, mode="1",
    remove_local_max=True, deblend=True
)

# Plotting result.
fig, ax = plt.subplots(1, 4, figsize=(24, 6))

# For keeping original and final images on same scale.
vmin = min(image.min(), galmasked.min())
vmax = max(image.max(), galmasked.max())

# fig.suptitle(filepath)
norm1 = ImageNormalize(image, vmin=vmin, vmax=vmax, interval=ZScaleInterval(), stretch=LogStretch())
im0 = ax[0].imshow(image, norm=norm1, origin='lower', cmap='gray')
ax[0].set_title("Original image")
cax0 = axes_colorbar(ax[0])
fig.colorbar(im0, cax=cax0, orientation='horizontal')

im1 = ax[1].imshow(orig_segmap, origin='lower')
ax[1].set_title("Original segmentation map (photutils)")
cax1 = axes_colorbar(ax[1])
fig.colorbar(im1, cax=cax1, orientation='horizontal')

im2 = ax[2].imshow(galsegmap, origin='lower', cmap='gray')
ax[2].set_title("Final segmentation map (galmask)")
cax2 = axes_colorbar(ax[2])
fig.colorbar(im2, cax=cax2, orientation='horizontal')

norm2 = ImageNormalize(galmasked, vmin=vmin, vmax=vmax, interval=ZScaleInterval(), stretch=LogStretch())
im3 = ax[3].imshow(galmasked, norm=norm2, origin='lower', cmap='gray')
ax[3].set_title("Final image (galmask)")
cax3 = axes_colorbar(ax[3])
fig.colorbar(im3, cax=cax3, orientation='horizontal')

plt.show()
```

Output:

![galmask_example](https://github.com/Yash-10/galmask/blob/main/example/galmask_example1.png)

> **_NOTE:_**  `orig_segmap` is the original segmentation map - it is not returned by galmask. It is an intermediate result calculated inside galmask (if a pre-calculated segmentation map is not input). Here the original segmentation map was stored in a FITS file for demonstration purposes. So if you pass `seg_image=None` (as done in the above example) and would like to create such four-column plots, you would need to edit the source code of `galmask.py` to save the internally calculated segmentation map in a FITS file.

# Documentation

The documentation is generated using the [Sphinx](https://www.sphinx-doc.org/) documentation tool and hosted by [Read the Docs](https://readthedocs.org/).
You can find the API reference and also some empirical tips to use galmask in the [documentation](https://galmask.readthedocs.io/en/latest/).

# Tests

For running the tests, you would need to install [pytest](https://docs.pytest.org/). You can navigate to the `tests/` directory and run:

```
pytest <name_of_file>
```

# Contribute

Contributions are welcome! Currently, there seem to be a few inefficient ways of handling things within galmask, and we would like you to contribute and improve the package!

Please let us know of any bugs/issues by opening an issue in the [issue tracker](https://github.com/Yash-10/galmask/issues).

# Citing

If you use `galmask` in your research, please consider citing the paper associated with this package:

```bibtex
@article{Gondhalekar_2022,
    doi = {10.3847/2515-5172/ac780b},
    url = {https://dx.doi.org/10.3847/2515-5172/ac780b},
    year = {2022},
    month = {jun},
    publisher = {The American Astronomical Society},
    volume = {6},
    number = {6},
    pages = {128},
    author = {Yash Gondhalekar and Rafael S. de Souza and Ana L. Chies-Santos},
    title = {galmask: A Python Package for Unsupervised Galaxy Masking},
    journal = {Research Notes of the AAS},
    abstract = {Galaxy morphological classification is a fundamental aspect of galaxy formation and evolution studies. Various machine learning tools have been developed for automated pipeline analysis of large-scale surveys, enabling a fast search for objects of interest. However, crowded regions in the image may pose a challenge as they can lead to bias in the learning algorithm. In this Research Note, we present galmask, an open-source package for unsupervised galaxy masking to isolate the central object of interest in the image. galmask is written in Python and can be installed from PyPI via the pip command.}
}
```

# License and copyright

galmask is licensed under the [MIT License](LICENSE).

galmask

galmask Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

galmask

galmask Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Download README