Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
10118193 | CATENA | 2019 | 9 Pages |
Abstract
Particle tracking velocimetry (PTV) is a promising image-based approach for remote streamflow measurements in natural environments. However, most PTV approaches require highly-defined round-shaped tracers, which are often difficult to observe outdoors. PTV-Stream offers a versatile alternative to cross-correlation-based PTV by affording the identification and tracking of features of any shape transiting in the field of view. This nearest-neighbor algorithm is inherently thought for estimating surface flow velocity of streams in outdoor conditions. The procedure allows for reconstructing and filtering the trajectories of features that are more likely to pertain to actual objects transiting in the field of view rather than to water reflections. The procedure is computationally efficient and is demonstrated to yield accurate measurements even in case of downsampled image sequences.
Keywords
Related Topics
Physical Sciences and Engineering
Earth and Planetary Sciences
Earth-Surface Processes
Authors
Flavia Tauro, Rodolfo Piscopia, Salvatore Grimaldi,