WorldFloods extended data set
- 1. Image Processing Laboratory, University of Valencia, Valencia, Spain
- 2. Image Processing Laboratory, University of Valencia, Valencia, Spain and Trillium Technologies, London, United Kingdom
- 3. School of Computer Science and Engineering, University of New South Wales (UNSW), Sydney, Australia
Description
"Global Flood Extent Segmentation in Optical Satellite Images" article data.
This data set is the extended version of the WorldFloods dataset released by Mateo-Garcia et al. (2021). We filtered low-quality floodmaps, extended the period of coverage to include flood events up to 2023, and manually fixed the labels of several flood maps. The resulting dataset has 509 flood extent maps from 144 different flood events.
The flood extent masks were visually inspected, and manually corrected when necessary, in order to provide reliable data to train supervised ML algorithms for flood extent segmentation. Here we provide the floodmaps.zip with vectorized reference masks, containing polygons of flood water, permanent water, clouds, and area of interest for each flood map.
Additionally, the metadatas.zip contains all the necessary information to download corresponding Sentinel-2 images, as well as the location of each flood event and activation code (according to Copernicus EMS, UNOSAT, or GLOFMIR conventions).
Portalés-Julià, E., Mateo-García, G., Purcell, C., & Gómez-Chova, L. Global flood extent segmentation in optical satellite images. Sci Rep 13, 20316 (2023). https://doi.org/10.1038/s41598-023-47595-7
This dataset is released under a Creative Commons non-commercial license (https://creativecommons.org/licenses/by-nc/4.0/legalcode.txt)
The development of this dataset has been supported by the Spanish Ministry of Science and Innovation project PID2019-109026RB-I00 (MINECO-ERDF MCIN/AEI/10.13039/501100011033).