Landsat-8 to Proba-V Transfer Learning and Domain Adaptation for Cloud detection

This repo contains the code of publication:

[1] G. Mateo-Garcia, V. Laparra, D. Lopez-Puigdollers, and L. Gomez-Chova, “Cross-Sensor Adversarial Domain Adaptation of Landsat-8 and Proba-V images for Cloud Detection,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, pp. 1–1, 2020, doi: 10.1109/JSTARS.2020.3031741. paper

Additional results can be found at https://isp.uv.es/projects/cloudsat/pvl8dagans

Requirements

The following code creates a new conda virtual environment with required dependencies.

conda create -n pvl8 -c conda-forge python=3.7 matplotlib 'pillow<7' numpy scipy libgdal=2.3 pandas tensorflow=2 --y

conda activate pvl8

pip install rasterio tifffile spectral tqdm luigi 

pip install h5py --ignore-installed --no-deps # To use pip h5py otherwise Proba-V images with szip compression cannot be read

Train proposed Adversarial Domain Adaptation

First, download the Biome Proba-V pseudo-simultaneous dataset from here. Then run:

python main_cycle_gans_da.py TrainCycleGAN --dataset-folder /folder/with/hdf5/file/ --suffix full --folder /path/to/save/models

Inference Adversarial Domain Adaptation

Apply the DA transformation [2] to a new Proba-V image. A level 2A 333M Proba-V image is required, it can be downloaded from the VITO portal. By default it uses the pretrained models from the checkpoints folder. See --cloud-detection-weights and --dagans-weights attributes.

python inference_da.py InferenceDA --pvimage PROBAV_L2A_20140321_150455_3_333M_V101.HDF5 --image-dest PROBAV_L2A_20140321_150455_3_333M_V101_ADAPTED.HDF5

The generated product contains the modified Proba-V bands and the cloud mask stored in CM_DAGANS dataset of the HDF5 file.

Transform Landsat-8 image to Proba-V following the physically based transformation

This is the Landsat-8 to Proba-V transformation first proposed in:

[2] G. Mateo-García, V. Laparra, D. López-Puigdollers, and L. Gómez-Chova, "Transferring deep learning models for cloud detection between Landsat-8 and Proba-V", ISPRS Journal of Photogrammetry and Remote Sensing, vol. 160, pp. 1–17, Feb. 2020, doi: 10.1016/j.isprsjprs.2019.11.024. paper

It requires a L1T Landsat-8 image that could be downloaded from the EarthExplorer. The --l8img attribute points to the unzipped folder with TIF images for each band.

python convert_landsat_probav.py ConvertPV --l8img LC08_L1TP_002054_20160520_20170324_01_T1 --outfolder folder/save/product/

Additionally it also accepts an image with its manually annotated cloud mask from the Biome dataset or from the 38-clouds dataset.

python convert_landsat_probav.py ConvertPV --l8img BC/LC80010112014080LGN00 --type-product biome --outfolder folder/save/product/

Cite

If you use this work please cite:

@article{mateo-garcia_cross-sensor_2020,
	title = {Cross-{Sensor} {Adversarial} {Domain} {Adaptation} of {Landsat}-8 and {Proba}-{V} images for {Cloud} {Detection}},
	issn = {2151-1535},
	doi = {10.1109/JSTARS.2020.3031741},
	journal = {IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing},
	author = {Mateo-Garcia, G. and Laparra, V. and Lopez-Puigdollers, D. and Gomez-Chova, L.},
	year = {2020},
	pages = {1--1},
}
 @article{mateo-garcia_transferring_2020,
	title = {Transferring deep learning models for cloud detection between {Landsat}-8 and {Proba}-{V}},
	volume = {160},
	issn = {0924-2716},
	doi = {10.1016/j.isprsjprs.2019.11.024},
	journal = {ISPRS Journal of Photogrammetry and Remote Sensing},
	author = {Mateo-García, Gonzalo and Laparra, Valero and López-Puigdollers, Dan and Gómez-Chova, Luis},
	month = feb,
	year = {2020},
	pages = {1--17},
}

Related work

• Benchmarking Deep Learning Models for Cloud Detection in Landsat-8 and Sentinel-2
• Multitemporal cloud masking in the Google Earth Engine

Acknowledgements

This work has been developed in the framework of the projects TEC2016-77741-R and PID2019-109026RB-I00 (MINECO-ERDF) granted to Luis Gómez-Chova.