| Résumé | Over the past decade, the use of imaging devices to perform quantitative measurements has seen wide-scale adoption and has become integral to the emerging fields of research, such as computer vision and artificial intelligence. Recent studies, published across a wide variety of fields, have demonstrated a vast number of ways through which image-based measurement systems can be used in their respective fields. A growing number of studies have demonstrated applications in coastal and ocean research. Edge detection methods have been used to measure water surface and bedform elevation from recorded video taken during wave flume experiments. The turbulent mixing of air and water, induced by the breaking waves and the runup processes, poses a particular problem for the edge-detection methods, since they rely on a sharp contrast between air and water. In this paper, an alternative method for tracking water surface, based on color segmentation, is presented. A set of experiments were conducted whereby the proposed method was used to detect water surface profiles for various types of breaking waves interacting with a rubble mound breakwater. The results were further processed to compute the surface velocity during runup. The time-history of surface velocity is shown to closely parallel the point measurements taken nearby the instrumented armor unit. These velocities can potentially serve as boundary conditions for determining the dynamic loads exerted on the armour units. Further, the image processing results are used to remove the time-varying buoyant force from the measured force acting on an individual armour unit, providing additional insight into how the forces develop over time. |
|---|
