GEMAUT : Un Outil pour Générer Automatiquement des Modèles Numériques de Terrain Haute Résolution à partir de Modèles Numériques de Surface satellitaires

Authors

  • Nicolas Champion IGN

DOI:

https://doi.org/10.52638/rfpt.2026.739

Keywords:

Digital Terrain Model, Digital Surface Model, Satellite, Optimisation, Open Source, Automatic Quality Assessment

Abstract

GEMAUT is a robust and open source tool for automatically generating Digital Terrain Models (DTMs) from high-resolution
satellite Digital Surface Models (DSMs). It performs fully automatic ground extraction - using either the SAGA slope-based filter
or the PDAL Cloth Simulation Filter - and reconstructs the terrain through a robust energy-based optimization scheme, without
requiring any external ground mask. A key contribution of this paper is the automatic quality-assessment module that derives
a spatialized estimate of relative vertical uncertainty from local DSM-DTM elevation differences, enabling reliable quality control
in operational workflows. GEMAUT has been entirely refactored using open-source libraries and publicly released on GitHub
to promote transparency, reproducibility, and collaboration within the SFPT community.

Downloads

Download data is not yet available.

Published

2026-06-08

How to Cite

Champion, N. (2026). GEMAUT : Un Outil pour Générer Automatiquement des Modèles Numériques de Terrain Haute Résolution à partir de Modèles Numériques de Surface satellitaires. Revue Française de Photogrammétrie et de Télédétection, 228(1), 33–42. https://doi.org/10.52638/rfpt.2026.739

Issue

Section

Special Issue: Data, algorithms and 3D applications

Categories