Fototex results

FOTOTEX: An Unsupervised Method for Urban Areas

The French UMR Espace-dev has just published an article entitled “Fast Unsupervised Multi-Scale Characterization of Urban Landscapes Based on Earth Observation Data” in the journal Remote Sensing, presenting the use of an algorithm dedicated to the study of urban areas by remote sensing based on texture information.

Three nested scales

With FOTOTEX, this paper introduces a unique, fast and unsupervised method to characterize urban areas at three nested scales: macro-scale (urban footprint), meso-scale (“neighbourhoods”) and micro-scale (urban objects). FOTOTEX combines Fourier’s Transform and Principal Component Analysis to convert the spatial signal (patterns in the image) into a frequency signal (number of times the pattern is repeated in a defined area). [Figure 1]

[Figure 1] The Fototex method enables the characterization of urban areas at three nested scales.
Figure 2. Résultats issus de l'algorithme FOTOTEX sur la ville de
Bouaké (Côte d’Ivoire) à trois échelles différentes : (a) délimitation de la tâche urbaine (b) identification d'unités urbaines, (c) densité urbaine et (d) extraction des contours d'objets urbains.
[Figure 2] Results from the FOTOTEX algorithm for the city of Bouaké (Ivory Coast) at three different scales: (a) delimitation of the urban footprint, (b) identification of urban units, (c) urban density and (d) extraction of urban object contours.

The results on Bouaké (Ivory Coast) [Figure 2] and Brasília (Brazil) have shown that a single Sentinel-2 image delineates the urban footprint (macro scale) better than some global products based on time series. Pleiades images identify different neighbourhoods and urban objects (meso- and micro-scales) using the texture signal, which is correlated with parameters such as building density, proportions of buildings and vegetation.

Research work based on FOTOTEX has been carried out in the framework of the TOSCA CNES APUREZA and DELICIOSA projects and is linked to the activities of the UMR Espace-Dev in the framework of Theia Risks Associated with Infectious Disease Risks SEC and Urban SEC.


Thibault Catry

Thibault Catry
IRD | Espace DEV
Contributions FR

Risks Associated with Infectious Disease Risks SEC | Urban SEC | FOTOTEX


Teillet, C.; Pillot, B.; Catry, T.; Demagistri, L.; Lyszczarz, D.; Lang, M.; Couteron, P.; Barbier, N.; Adou Kouassi, A.; Gunther, Q. ; et al. Fast Unsupervised Multi-Scale Characterization of Urban Landscapes Based on Earth Observation Data. Remote Sens. 2021,13, 2398.

Read THEIA Bulletin n°15

>>> Dowload PDF version
>>> Read on Calaméo

>>> Read older issues

More news

clASpy_T : a 3D point cloud classification tool for estuary monitoring

The THEIA portfolio of Lidar data processing tools has been extended with a new software tool for creating 3D point cloud classification models using machine learning: cLASpy_T cLASpy_T aims to […]

Two THEIA LiDAR tools for forest monitoring

On this International Day of Forests, for which innovation has a major role to play in 2024, we are highlighting two tools dedicated to the management of aerial LiDAR data […]

SWOT data that exceeds expectations

The SWOT space mission was launched at the end of 2022 by CNES and NASA to document the water masses of the entire planet – oceans and continental waters. Data […]