Skip to content

icey-zhang/DIAnet

Repository files navigation

Distribution-aware Interactive Attention Network and Large-scale Cloud Recognition Benchmark on FY-4A Satellite Image

Paper

⭐ our article

Citation

@article{zhang2024distribution,
  title={Distribution-aware Interactive Attention Network and Large-scale Cloud Recognition Benchmark on FY-4A Satellite Image},
  author={Zhang, Jiaqing and Lei, Jie and Xie, Weiying and Jiang, Kai and Cao, Mingxiang and Li, Yunsong},
  journal={arXiv preprint arXiv:2401.03182},
  year={2024}
}

Products

Usage of the code for FYH dataset

Prepare the dataset

Download the FY4A L1 dataset for FY4A

Download the Himawari dataset for Himawari

├── fydatahimawari
│   ├── Himawari
......
│   │   ├── 202005
│   │   │   ├── 05
│   │   │   │   ├── 05
│   │   │   │   │   ├──NC_H08_20200505_0500_L2CLP010_FLDK.02401_02401.nc
│   │   │   │   │   ├──NC_H08_20200505_0510_L2CLP010_FLDK.02401_02401.nc
│   │   │   │   │   ├──NC_H08_20200505_0520_L2CLP010_FLDK.02401_02401.nc
......

│   ├── FY4A
......

│   │   ├── 20200104
│   │   │   ├── FY4A-_AGRI--_N_REGC_1047E_L1-_FDI-_MULT_NOM_20200104003000_20200104003417_4000M_V0001.HDF
│   │   │   ├── FY4A-_AGRI--_N_REGC_1047E_L1-_FDI-_MULT_NOM_20200104003418_20200104003835_4000M_V0001.HDF
......

Split the dataset

python split.py

Train the model

python tools/train.py --cfg experiments/cityscapes/seg_hrnet_AWCA_PSNL_z_w48_train_100x100_sgd_lr1e-2_wd5e-4_bs_8_epoch100.yaml

Test the production

python tools/test_production.py

The generation validation of the SPARCS dataset

Prepare the dataset

Download the Cloud detection SPARCS dataset

Split the dataset

python split_sparcs.py

Train the model

python tools/train.py  --cfg experiments/cityscapes/seg_hrnet_AWCA_PSNL_z_w48_train_200x200_sgd_lr1e-4_wd5e-4_bs_6_epoch100_sparcs.yaml

Test the model

python tools/test.py  --cfg experiments/cityscapes/seg_hrnet_AWCA_PSNL_z_w48_train_200x200_sgd_lr1e-4_wd5e-4_bs_6_epoch100_sparcs.yaml

Gernerate the final result

python tools/test_pic.py --cfg experiments/cityscapes/seg_hrnet_AWCA_PSNL_z_w48_train_200x200_sgd_lr1e-4_wd5e-4_bs_6_epoch100_sparcs.yaml
python tools/test_pic_merge.py

About

DIAnet is accepted by TGRS

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published