Reference : Image processing techniques for velocity estimation in highly aerated flows: Bubble I...
Scientific congresses and symposiums : Paper published in a book
Engineering, computing & technology : Civil engineering
Image processing techniques for velocity estimation in highly aerated flows: Bubble Image Velocimetry vs. Optical Flow
Bung, Daniel B. mailto []
Valero Huerta, Daniel mailto [Université de Liège - ULiège > > > Form. doct. sc. ingé. & techn. (archi., gén. civ. - paysage)]
Sustainable Hydraulics in the Era of Global Change
Erpicum, Sébastien mailto
Dewals, Benjamin mailto
Archambeau, Pierre mailto
Pirotton, Michel mailto
Taylor & Francis Group
4th IAHR Europe Congress
from 27-7-2016 to 29-7-2016
University of Liège
[en] Measuring of flow velocities in aerated flows is known to be difficult in physical models. Application of classical anemometers or ADV probes is limited to low air concentration. Thus, highly aerated flows are commonly investigated by use of intrusive needle probes (conductivity or optical fiber) which allow determination of both, air concentration and velocity, as well as related parameters (e.g. bubble chord lengths and turbulence). In the recent past, non-intrusive image processing techniques have gained more attraction. In the present paper, Bubble Image Velocimetry and Optical Flow methods are applied to aerated stepped spillway flows using high-speed cameras with different resolution to highlight capabilities and limitations of both methods.
Results show that Optical Flow is capable to give results of at least the same accuracy as Bubble Image Velocimetry. A higher image resolution enhances the quality of the results. The dense velocity information being obtained by Optical Flow may help to carry out investigations on turbulence in future.
Urban and Environmental Engineering

File(s) associated to this reference

Fulltext file(s):

Restricted access
ch028.pdfPublisher postprint3.67 MBRequest copy

Bookmark and Share SFX Query

All documents in ORBi are protected by a user license.