Towards a detection and recognition system for freshwater fish

Broersen, Jeroen (2009) Towards a detection and recognition system for freshwater fish.

Abstract:FishFlow Innovations is a company that invents and produces among other things fish passages. To determine the effectiveness of such passages, it is necessary to continually count and recognize passing fish. Up to now, counting and classification of fish is mainly done by using a fyke net. Because this process is very labour-intensive, possibilities for electronic counting and recognition are investigated. In this project, various methods for automatically counting and recognising passing fish has been studied, such as sonar, vision, impedance measurements, optical fish detection and temperature measurements. Especially optical and electrical impedance measurement methods appeared to have good prospectives and have, therefore, been investigated in more detail. Two electrical measurement methods have been studied in detail: capacitive and resistive. Simulations and experiments showed that capacitive measurement are not feasible in practice. The resistive method, which works by using the difference in conductivity between fish and water proves to be useful to detect fish. Fish species recognition using impedance tomography is also studied. An experiment and simulations showed that detection of fish species by tomographic measurements is not feasible. Small details, such as prickles are almost impossible to detect using these measurements. However, the presence of a fish can be detected using this resistive method. For optical methods, especially the use of cameras is studied. In order to test these optical methods, an experiment is carried out in Enschede at the University. The goal of this experiment was to find out if it is possible to get enough detail for fish species recognition when looking through a layer of water. Furthermore a test setup is built in a real fish passage from FishFlow Innovations in Roermond. This setup has unfortunately produced little results in counted fish. However, this setup has brought some insight in the lighting, and the necessary frame rate. For further research, it is recommended to investigate different setups of light sources for the camera setup. It is also recommended to research the effects of water for more sources (e.g. other rivers). For tomographic methods, it is advised to test setups with more electrodes to get a higher resolution resistance map.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:53 electrotechnology
Programme:Electrical Engineering MSc (60353)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page