IAHR Document Library

« Back to Library Homepage « Proceedings of the 39th IAHR World Congress (Granada, 2022)

Video-Based Bedload Measurement in a Large River

Author(s): Alexander Anatol Ermilov; Sandor Baranya; Gabor Fleit; Slaven Conevski; Massimo Guerrero; Nils Ruther

Linked Author(s): Alexander Anatol Ermilov, Sándor Baranya, Gábor Fleit, Slaven Conevski, Massimo Guerrero

Keywords: Bedload; Image processing; Video-based; Hydromorphology

Abstract: The river flow and the sediment that composes the riverbed are in constant interaction. The former, depending on its energy, can erode the latter and depose it somewhere else, while the riverbed poses hydraulic resistance. Despite its importance, this interaction still has undiscovered aspects and unanswered questions. In alluvial rivers, one important factor that describes its morphological changes is the bedload yield. Defining its value however, remains a complex task and as a result of that, solely theoretical approaches are not sufficient. Hence, the emphasis from laboratory measurements and empirical methods seems to be shifting towards the actual observation of nature on the field. In spite of this, the traditional sampling methods are biased and also energy consuming. The behaviour of the current sample instruments happen to be arbitrary and unreliable. For instance, the sampler can easily grab into the riverbed at the end of the measurement, filling itself with bedmaterial and adding its mass inseparably as an error to our result. This and other unreliabilities led to the need of new ways of bedload measurement. Nowadays, computer vision and image processing are swiftly spreading across all type of fields of research, replacing older measurement methods or opening up new aspects of observations. The field of hydromorphology is no exception. In the case of bedload, by applying underwater cameras on the traditional sampling instruments, we can observe their behaviour and adjust them accordingly. On the other hand, bedload yield may be estimated via certain image processing methods and further used to calibrate, or even replace, our regular measurements. The latter one also opens up the possibility of a faster and automatized bedload analysis. This paper explores the possibilities of applying such method in a large river (Danube), comparing video-based calculations to traditional ones. Our first set of results and experience will be presented along with further possibilities of improvements.


Year: 2022

Copyright © 2024 International Association for Hydro-Environment Engineering and Research. All rights reserved. | Terms and Conditions