Fishes are in risk of injury during their downstream migration via hydro power plants. The knowledge on fish behavior in the near field of fish guidance structures is of importance for both the evaluation of their effectiveness and their optimization. However, up to now, fish-observations have been only qualitatively realized. In the first part of this project an automatized Fish-Tracking Module had been developed that is able to detect either the path of a single fish or the midpoint of a schooling fish from video recordings obtained with etho-hydraulic laboratory flume experiments. The module is mainly based on the Computer Vision System Toolbox™ of the commercial software MATLAB. However, an unambiguous separation of the full length of each fish in a schooling still remains challenging. In the second part of the project, the Fish-Tracking Module has now successfully been automatized, further developed, and applied to a large data set of different fish species. The newly developed tool is now ready to be applied for analyzing large data sets.
Auftragnehmer/Contractant/Contraente/Contractor:
ETH Zürich, Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie (VAW)
Autorschaft/Auteurs/Autori/Authors:
Detert,Martin
Albayrak,Ismail
Boes,Robert