This is the source code for our paper: Hyperspectral image denoising via spectral noise distribution bootstrap.
- Python =3.7
- torch =1.9.0, torchnet, torchvision
- pickle, tqdm, tensorboardX, scikit-image
-
download ICVL hyperspectral image database from here
save the data in *.mat format into your folder
-
generate data with synthetic noise for training and validation
# change the data folder first python ./data/datacreate.py
python main.py -a nnet --dataroot (your data root) --phase train
-
Testing on Synthetic data or Real HSIs with the pre-trained model
python main.py -a nnet --phase test -r -rp checkpoints/model_best.pth
If you find this work useful, please cite our paper:
@article{Pan2023hypersepctral,
title = {Hyperspectral image denoising via spectral noise distribution bootstrap},
author = {Erting Pan and Yong Ma and Xiaoguang Mei and Fan Fan and Jiayi Ma},
journal = {Pattern Recognition},
volume = {142},
pages = {109699},
year = {2023},
issn = {0031-3203},
doi = {https://doi.org/10.1016/j.patcog.2023.109699}
}
Feel free to open an issue if you have any question. You could also directly contact us through email at panerting@whu.edu.cn (Erting Pan)