Code for this paper.
Joint Spatial-Spectral Pattern Optimization and Image Reconstruction for Hyperspectral Imaging (IEEE-JSTSP 2022)
Tao Zhang, Zhiyuan Liang and Ying Fu.
In this paper, we propose a deep-learning based method for high quality hyperspectral imaging by joint spatial-spectral optimization, including joint multispectral filter array (MSFA) and spectral sensitivity function (SSF) optimization, joint spatial demosaicing (SpaDM) and spectral super-resolution (SpeSR), and joint pattern optimization and HSI reconstruction.
- Python >= 3.6, PyTorch >= 1.7.1
conda install -c conda-forge python-lmdb
conda install caffe
pip install --upgrade git+https://github.com/pytorch/tnt.git@master
pip install -r requirements.txt
To generate the taining and testing data, run
python utility/lmdb_data.py
python hsi_dm.py -a admmn_16channel_alpha -d /Path/to/Your/Data
-
Selection
python hsi_dm.py -a admmn_alpha -d /Path/to/Your/Data
-
Design
python hsi_dm.py -a admmn_spectrum -d /Path/to/Your/Data
python hsi_test.py -a admmn_16channel_alpha --use-2dconv -r -rp /Path/to/Your/Model
If you find this work useful for your research, please cite:
@ARTICLE{zhang2022joint,
author={Zhang, Tao and Liang, Zhiyuan and Fu, Ying},
journal={IEEE Journal of Selected Topics in Signal Processing},
title={Joint Spatial-Spectral Pattern Optimization and Hyperspectral Image Reconstruction},
year={2022},
volume={16},
number={4},
pages={636-648},
}