Articles | Volume 14, issue 4
https://doi.org/10.5194/esurf-14-527-2026
© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/esurf-14-527-2026
© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
14 Jul 2026
Research article |
| 14 Jul 2026
| 14 Jul 2026
ImageGrains 2.0: Improved precision and generalization for grain segmentation
Institute of Geological Sciences, University of Bern, Bern, 3012, Switzerland
Guillaume Witz
Data Science Lab, University of Bern, Bern, 3012, Switzerland
Ariel Do Prado
Institute of Geological Sciences, University of Bern, Bern, 3012, Switzerland
Institute of Geosciences, University of São Paulo, São Paulo, 05508-080, Brazil
Philippos Garefalakis
Institute of Geological Sciences, University of Bern, Bern, 3012, Switzerland
Amanda Wild
Institute of Physical Geography and Geoecology, RWTH-Aachen University, 52062 Aachen, Germany
GFZ Helmholtz Centre for Geosciences, 14473 Potsdam, Germany
Fanny Ville
Fluvial Dynamics Research Group (RIUS), University of Lleida, Lleida, 25003, Spain
Bennet Schuster
Institute of Geological Sciences, University of Bern, Bern, 3012, Switzerland
Michael Horn
Data Science Lab, University of Bern, Bern, 3012, Switzerland
Jürgen Österle
School of Geography, Environment and Earth Sciences, Victoria University of Wellington, Wellington, 6012, New Zealand
Amt der Vorarlberger Landesregierung, Bregenz, 6901, Austria
Stefano C. Fabbri
Federal Office of Topography swisstopo, Wabern, 3084, Switzerland
Camille Litty
Federal Office of Topography swisstopo, Wabern, 3084, Switzerland
Stefan Achleitner
Unit of Hydraulic Engineering, University of Innsbruck, Innsbruck, 6020, Austria
Sebastian Leistner
Unit of Hydraulic Engineering, University of Innsbruck, Innsbruck, 6020, Austria
Clemens Hiller
Unit of Hydraulic Engineering, University of Innsbruck, Innsbruck, 6020, Austria
Natural Hazards and Risk Management, Geoconsult ZT GmbH, Puch bei Hallein, 5412, Austria
Fritz Schlunegger
Institute of Geological Sciences, University of Bern, Bern, 3012, Switzerland
Data sets
ImageGrains 2.0 dataset D. Mair, A. do Prado, P. Garefalakis, et al. https://doi.org/10.5281/zenodo.17866827
Model code and software
ImageGrains 2.0 models D. Mair, G. Witz, A. do Prado, A., et al. https://doi.org/10.5281/zenodo.15309323
Short summary
This study introduces an updated image analysis framework for automatically identifying and measuring sediment grains in various types of images and scans. We employ a high-performing segmentation approach for a wide range of geoscientific data, using carefully curated ground truth data. The method achieves higher accuracy and more consistent measurements than existing approaches. The data and algorithm are openly available and provided in a user-friendly way.
This study introduces an updated image analysis framework for automatically identifying and...
