!!! Repository under preparation !!!

We uploaded code and dataset in a raw format for pubblication purposes

The code is splitted in multiple Jupyter Notebooks, we are currently refining it to make the code cleaner and more readable.

The dataset is splitted in multiple folder, we are currently fixing the dataset structure to fit the description in the paper

Spatio-Temporal SAR-Optical Data Fusion for Cloud Removal via a Deep Hierarchical Model

Authors: Alessandro Sebastianelli, Erika Puglisi, Maria Pia Del Rosso, Jamila Mifdal, Artur Nowakowki, Fiora Pirri, Pierre Philippe Mathieu and Silvia Liberata Ullo

Dataset:

The dataset collects roughly 8000 S1 and 8000 S2 images or 2000 S1 and 2000 S2 time-series of 4 images. Each image has a shape of 256 X 256 pixels; for the S1 images only the VV polarisation has been considered, resulting in a matrix with a shape of 256 X 256 X 1 ; for the S2 the red, green and blue bands have been considered, resulting in a matrix with a shape of 256 X 256 X 3, and an extra matrix with a shape of 256 X 256 X 1 for the cloud mask (QA60 band).

The dataset has been created using our tool proposed in: Sebastianelli, A., Del Rosso, M. P., & Ullo, S. L. (2021). Automatic dataset builder for Machine Learning applications to satellite imagery. SoftwareX, 15, 100739.

Part 1:

Part 2:

https://drive.google.com/drive/folders/1Af_V8uY-OAtW4O_L_doSlPsmpueZdd11?usp=sharing

Papers

Sebastianelli, A., Nowakowski, A., Puglisi, E., Del Rosso, M. P., Mifdal, J., Pirri, F., ... & Ullo, S. L. (2021). Sentinel-1 and Sentinel-2 Spatio-Temporal Data Fusion for Clouds Removal. arXiv preprint arXiv:2106.12226. https://arxiv.org/abs/2106.12226