Articles | Volume 11, issue 6
Research article
01 Nov 2023
Research article |  | 01 Nov 2023

Automated riverbed composition analysis using deep learning on underwater images

Alexander A. Ermilov, Gergely Benkő, and Sándor Baranya

Related subject area

Physical: Geomorphology (including all aspects of fluvial, coastal, aeolian, hillslope and glacial geomorphology)
On the use of packing models for the prediction of fluvial sediment porosity
Christoph Rettinger, Mina Tabesh, Ulrich Rüde, Stefan Vollmer, and Roy M. Frings
Earth Surf. Dynam., 11, 1097–1115,,, 2023
Short summary
Marsh-induced backwater: the influence of non-fluvial sedimentation on a delta's channel morphology and kinematics
Kelly M. Sanks, John B. Shaw, Samuel M. Zapp, José Silvestre, Ripul Dutt, and Kyle M. Straub
Earth Surf. Dynam., 11, 1035–1060,,, 2023
Short summary
Spatial and temporal variations in rockwall erosion rates derived from cosmogenic 10Be in medial moraines at five valley glaciers around Pigne d'Arolla, Switzerland
Katharina Wetterauer and Dirk Scherler
Earth Surf. Dynam., 11, 1013–1033,,, 2023
Short summary
Building a bimodal landscape: bedrock lithology and bed thickness controls on the morphology of Last Chance Canyon, New Mexico, USA
Sam Anderson, Nicole Gasparini, and Joel Johnson
Earth Surf. Dynam., 11, 995–1011,,, 2023
Short summary
Geotechnical controls on erodibility in fluvial impact erosion
Jens M. Turowski, Gunnar Pruß, Anne Voigtländer, Andreas Ludwig, Angela Landgraf, Florian Kober, and Audrey Bonnelye
Earth Surf. Dynam., 11, 979–994,,, 2023
Short summary

Cited articles

Adams, J.: Gravel Size Analysis from Photographs, J. Hydraul. Div., 1979, 105, 1247–1255,, 1979. 
Baranya, S., Fleit, G., Józsa, J., Szalóky, Z., Tóth, B., Czeglédi, I., and Erős, T.: Habitat mapping of riverine fish by means of hydromorphological tools, Ecohydrology, 11, e2009,, 2018. 
Barnard, P., Rubin, D., Harney, J., and Mustain, N.: Field test comparison of an autocorrelation technique for determining grain size using a digital beachball camera versus traditional methods, Sediment. Geol., 201, 180–195, 2007. 
Benjankar, R., Tonina, D., and Mckean, J.: One-dimensional and two-dimensional hydrodynamic modelling derived flow properties: Impacts on aquatic habitat quality predictions, Earth Surf. Proc. Land., 40, 340–356, 2015. 
Benkő, G., Baranya, S., Török, T. G., and Molnár, B.: Folyami mederanyag szemösszetételének vizsgálata Mély Tanulás eljárással drónfelvételek alapján (in English: Analysis of composition of riverbed material with Deep Learning based on drone video footages), Hidrológiai Közlöny, 100, 61–69, 2020. 

The requested paper has a corresponding corrigendum published. Please read the corrigendum first before downloading the article.

Short summary
A novel, artificial-intelligence-based riverbed sediment analysis methodology is introduced that uses underwater images to identify the characteristic sediment classes. The main novelties of the procedure are as follows: underwater images are used, the method enables continuous mapping of the riverbed along the measurement vessel’s route contrary to conventional techniques, the method is cost-efficient, and the method works without scaling.