Bedforms are common morphological features that develop in many riverine environments when flow conditions are sufficient to initiate bedload transport. They play a relevant role in the development of several hydraulic and morphological processes, like flow resistance and sediment transport. At the same time, bedforms can hinder navigation and threaten the structural integrity of hydraulic structures. Understanding their spatiotemporal evolution is therefore essential for effective river management, roughness estimation and flooding predictions, sediment budget assessment, and dredging projects. Traditional bedform analysis and characterization mainly focus on longitudinal (one-dimensional) sections of the river bed. Traditional approaches for analyzing single bedforms are the so-called Bedform Tracking Methods. While relatively simple to implement, 1D bedform tracking algorithms may miss the identification of large scale bedforms when smaller features are present. Lee et al. (2021) used two-dimensional frequency–wavenumber spectral analysis to estimate scale-dependent bedform migration velocities. They showed that Fourier-based methods effectively capture large- and medium-scale bedform dynamics. However, this method requires scale separation to resolve slowly migrating features masked by smaller bedforms. In contrast, direct 2D analysis of spatio-temporal bathymetric evolution may open new avenues for investigating bedform dynamics. These approaches enable the tracking of multidirectional characteristics, deformation processes (merging versus splitting), and bidimensional migration velocities. In this context, image processing algorithms like Bathymetric Image Velocimetry (BIV) already exist and allow 2D cross-correlation analyses to track bedform front velocities. Our project aims to build on this expertise by developing an algorithm capable of performing a comprehensive direct 2D analysis.