OBIA Classification of Riverine Vegetation in a Small Open Channel Using RGB Drone Imagery

In temperate areas, small watercourses, especially agricultural ditches, are typically surrounded by seasonally changing vegetation, which significantly influences hydrodynamic and ecological processes within and around the channel.However, research on how different vegetation maintenance practices affect flow and mixing processes at the reach scale remains limited.Addressing this knowledge gap requires a series of field experiments conducted under varying flow and vegetation conditions, along with a simple and accessible method for vegetation characterisation.This study evaluates the efficiency of RGB drone imagery in mapping riverine vegetation using OBIA classifiers.For low vegetation coverage, SVM combined with Haralick textural features provided the best results, while for high vegetation, SVM combined with DEM delivered the best classification outcomes.