LRRSSN is a novel method for thick cloud removal in multitemporal remote sensing images (MRSIs). It integrates model-driven low-rank and sparse decomposition with a data-driven self-supervised deep network. The method iteratively estimates the clean image and cloud component using an HQS-based optimization framework and employs a Guided Deep Decoder (GDD) to capture deep priors without external training data. The overall framework is illustrated below:

If you use this method in your research, please cite:

@article{chen2024lrrssn,
  author={Chen, Yong and Chen, Maolin and He, Wei and Zeng, Jinshan and Huang, Min and Zheng, Yu-Bang},
  journal={IEEE Transactions on Geoscience and Remote Sensing}, 
  title={Thick Cloud Removal in Multitemporal Remote Sensing Images via Low-Rank Regularized Self-Supervised Network}, 
  year={2024},
  volume={62},
  number={},
  pages={1-13},
  doi={10.1109/TGRS.2024.3358493}}

📄 IEEE Link | DOI: 10.1109/TGRS.2024.3358493

1. You can install the dependencies with:

pip install -r requirements.txt

2. Clone the repository:

git clone https://github.com/try-agaaain/LRRSSN.git
cd LRRSSN

3. Run the main script:

python main.py

This project is licensed under the CC BY-NC-ND 4.0. The code is free to use for academic research and non-commercial purposes only.