Image processing and form recognition applied to the quantitative visualisation of coherent flow structures.
Flow visualisation methods such as dye tracers have long been a core methodology for the analysis of turbulent flows. These methods are ideally suited to qualitative observations of coherent structures and their past usage has yielded important insights into turbulent flows. However, the analysis of flow visualisation data need not be limited to qualitative observations. Digital image processing and basic form recognition methods largely developed in the context of remote sensing and earth observation can be applied to flow visualisation experiments in order to extract quantitative information. This paper will demonstrate how such methods can be used on digital films of dye tracer experiments. Specifically, we will examine naturally occurring flow structures observed during a dye tracer experiment conducted in a gravel bed river in Quebec, Canada. The image analysis will be applied in order to automatically identify individual coherent flow structures, measure their size, their orientation in the flow and finally their mean downstream velocity. This novel application of image processing methods to dye tracer experiments allows for quantitative flow visualisations which in turn yield a much more detailed description of coherent flow structures.
AGU Fall Meeting Abstracts
- Pub Date:
- December 2007
- 1625 Geomorphology and weathering (0790;
- 1855 Remote sensing (1640);
- 1856 River channels (0483;
- 1860 Streamflow;
- 1872 Time series analysis (3270;