The measurement of river-bed grain-sizes is a key parameter in many formulas for calculating the hydraulic characteristics
of a river and the associated solid flow. Sampling methods for measuring grain sizes were developed during the 20th
century and remain in common use. However, these traditional methods require extensive field work to obtain results that
are representative of the heterogeneity of the spatial distribution of grain sizes. Moreover, they do not normally provide
a continuous spatial representation of the grain size distribution. New automated image-based measurement methods
have been developed recently. These methods can provide valuable information on the spatial distribution of bed-surface
grain sizes on the scale of several hundred meters without having to perturb the granular bed structure. The application
of these automated digital approaches in the context of mountain rivers is of great interest, because of the diversity of
their bed-material grain sizes. These specific conditions could be very challenging for automated digital approaches.
Several automated measurement methods were used to analyze orthoimages and digital elevation models of an alpine
gravel-bed river, which were reconstructed from UAV collected images. A comparison between the results obtained
from the application of Fehr’s line sampling method and some automated methods was performed. Limitations of these
frameworks were identified that would need to be tackled in order to provide more accurate grain-size estimates. Based
on the information obtained by the automated methods, an analysis of the spatial distribution of sediments was performed.
It revealed how grain clusters tend to be distributed onto this mountain river bed.