Convolutional neural networks for global human settlements mapping from Sentinel-2 satellite imagery
- PDF / 22,627,887 Bytes
- 24 Pages / 595.276 x 790.866 pts Page_size
- 46 Downloads / 185 Views
(0123456789().,-volV)(0123456789(). ,- volV)
ORIGINAL ARTICLE
Convolutional neural networks for global human settlements mapping from Sentinel-2 satellite imagery Christina Corbane1 • Vasileios Syrris1 • Filip Sabo2 • Panagiotis Politis2 • Michele Melchiorri3 Martino Pesaresi1 • Pierre Soille1 • Thomas Kemper1
•
Received: 18 May 2020 / Accepted: 14 October 2020 The Author(s) 2020
Abstract Spatially consistent and up-to-date maps of human settlements are crucial for addressing policies related to urbanization and sustainability, especially in the era of an increasingly urbanized world. The availability of open and free Sentinel-2 data of the Copernicus Earth Observation program offers a new opportunity for wall-to-wall mapping of human settlements at a global scale. This paper presents a deep-learning-based framework for a fully automated extraction of built-up areas at a spatial resolution of 10 m from a global composite of Sentinel-2 imagery. A multi-neuro modeling methodology building on a simple Convolution Neural Networks architecture for pixel-wise image classification of built-up areas is developed. The core features of the proposed model are the image patch of size 5 9 5 pixels adequate for describing built-up areas from Sentinel-2 imagery and the lightweight topology with a total number of 1,448,578 trainable parameters and 4 2D convolutional layers and 2 flattened layers. The deployment of the model on the global Sentinel-2 image composite provides the most detailed and complete map reporting about built-up areas for reference year 2018. The validation of the results with an independent reference dataset of building footprints covering 277 sites across the world establishes the reliability of the built-up layer produced by the proposed framework and the model robustness. The results of this study contribute to cutting-edge research in the field of automated built-up areas mapping from remote sensing data and establish a new reference layer for the analysis of the spatial distribution of human settlements across the rural–urban continuum. Keywords Convolutional neural networks Remote sensing Image segmentation Human settlements Built-up areas
1 Introduction New ways to map and measure the built-up environment over large areas are critical to answering a wide range of research questions and to addressing policies related to urbanization and sustainability. This is particularly true in the era of an increasingly urbanized world [1]. Earth
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00521-020-05449-7) contains supplementary material, which is available to authorized users. & Christina Corbane [email protected] 1
European Commission, Joint Research Centre (JRC), Ispra, Italy
2
Arhs Developments S.A, 4370 Belvaux, Luxembourg
3
Engineering S.p.a, 00144 Rome, Italy
Observation (EO) has become a promising tool to provide up to date geospatial information on the status and dynamics of built-up areas and human settlements [2]. With
