CNN | Sentinel-2 | GHSL | Built-up

GHS-BUILT-S2 R2020A

GHS built-up grid, derived from Sentinel-2 global image composite for reference year 2018 using Convolutional Neural Networks (GHS-S2Net)

GHS-BUILT-S2 R2020A

built-up grid derived from Sentinel-2 global image composite for reference year 2018 using Convolutional Neural Networks (GHS-S2Net).

It builds on a new Deep Learning framework for pixel-wise large-scale classification of built-up areas named GHS-S2Net.

(GHS stands for Global Human Settlements, S2 refers to the Sentinel-2 satellite).

Download the GHS-BUILT-S2 R2020A dataset.

How to cite:

Dataset:
Corbane, Christina; Sabo, Filip; Politis, Panagiotis; Syrris Vasileos:
GHS-BUILT-S2 R2020A - built-up grid derived from Sentinel-2 global image composite for reference year 2018 using Convolutional Neural Networks (GHS-S2Net). European Commission, Joint Research Centre (JRC) (2020)
PID: http://data.europa.eu/89h/016d1a34-b184-42dc-b586-e10b915dd863 , doi:10.2905/016D1A34-B184-42DC-B586-E10B915DD863

Concept & Methodology:
Corbane, C., Syrris, V., Sabo, F. et al.
Convolutional neural networks for global human settlements mapping from Sentinel-2 satellite imagery. Neural Comput & Applic (2020).
doi:10.1007/s00521-020-05449-7

Product name:
GHS_BUILT_S2comp2018_GLOBE_R2020A
Coordinate system:
Each tile inherits the projection of the UTM grid zone to which it belongs to.
Check file index UTM_land.shp for the code of the UTM grid zones
Resolutions available:
10m
Data type:
8 bit
Description:
This dataset corresponds to global map of built-up areas expressed in terms of a probability grid at 10 m spatial resolution derived from a Sentinel-2 global image composite (GHS_composite_S2_L1C_2017-2018_GLOBE_R2020A_UTM_10_v1_0) for reference year 2018.
Dataset name:
GHS_BUILT_S2comp2018_GLOBE_R2020A
Legend:
Built-up probability values rescaled in the range 0-100 (0 for probability = 0 and 100 for probability = 1)
255 = nodata