Skip to content

Latest commit

 

History

History
15 lines (12 loc) · 832 Bytes

README.md

File metadata and controls

15 lines (12 loc) · 832 Bytes

AstroNet

AstroNet is a repository that makes simple to train an AI model in the realm of astrophotography.

It has been developed, in particular, for the GraXpert team and the fantastic work done by Steffen has been a fondamental inspiration for this library.

Model supported

Name Training Preprocessing
Background extraction train_bge.py data/prepare_fits_bge.py
Denoise train_denoise.py data/prepare_fits_denoise.py

Train background extraction

  1. python data/prepare_fits_bge.py --img_folder <images with gradients> --bkg_folder <gradients> --output_folder <folder in which you want the dataset to be created>
  2. python train_bge.py