underwater fish detector

[fedimakni]

The capacitance check is very interesting, but i think it require to use 2 plates exposed to sea water and then measure its capacitance correct?
i think it can work too, but i think it require a good material choice as the sea water can affect those plates over time right?

[eutectique]

The capacitance check is very interesting, but i think it require to use 2 plates exposed to sea water and then measure its capacitance correct?

You will then measure conductivity of sea water and fish within. You should avoid exposing the plates.

[fedimakni]

But if they are not exposed to sea water how the plates will act as capacitance? i think with around 20-30 cm gap it will be really not noticeable enough. What do you think?

[eutectique]

But if they are not exposed to sea water how the plates will act as capacitance?

Here is a good explanation of the theory behind the capacitor: https://en.wikipedia.org/wiki/Capacitor#Theory_of_operation

If you expose two plates to sea water they will form a resistor, generally speaking.

[Kleinstein]

One could have the electrodes in direct contact to the water or have a thin isolating layer on top. The isolator acts like a series capacitor and would reduce the current flow quite a bit.
Not to distract the fishes one would have to use a weak field anyway. With seewater the water is likely more in the resistive regime and little actual capacitance. The fish than would be more a change in resistance. The system would thus also react to chances in water composition (e.g. after rain or fresh water from a river).

[voltsandjolts]

We have a fish ladder locally with a fish counter in it. Uses a trained AI camera system apparently.

voltsandjolts,
Do you have a link for that product? thanks

No technical info, just some general info and press releases from SSE, the hydro dam owner.
https://www.pitlochrydam.com/
https://tayrivers.org/news/pitlochry-dam-fish-counter/

There is also this description of an AI fish detection system:
https://www.anthropocenemagazine.org/2023/10/artificial-intelligence-it-turns-out-excels-at-the-humble-pursuit-of-counting-fish/
https://blog.roboflow.com/using-computer-vision-to-count-fish-populations/

[Kleinstein]

The camera + AI adds the identification of the fishes. For just counting it would be more overkill.

[qtbits]

Personally - I'd go with the ESP32 + camera - There are already examples for "face" recognition and with a bit of effort it could be "trained" for fish. Also... you may want to do more than just counting a a later date.
Plus you get the wireless (shouldn't "fry" the fish too much ).

[Njk]

Remember the air flow sensor in certain car engines. There is a filament (or film) resistor which is heated by current and cooled by air flow. The flow is measured by measurement the resistance value that depends on the temperature. Use a setup like this. When the fish is passing through the tunnel, it changes the thermal balance inside the tunnel. To simplify the detection, use a set of two similar tunnels and measure the difference (assuming social interaction in fish society is not so advanced so they will not try to fool you by passing both the tunnels simultaneously).

[Harrow]

But if they are not exposed to sea water how the plates will act as capacitance? i think with around 20-30 cm gap it will be really not noticeable enough. What do you think?

I'm not sure why, but I was visualising a large salmon squeezing through a 4-inch tube. With a larger tube and smaller fish, I'm guessing it's probably not noticeable. Now I wish I had some salmon and a tube to find out. I also wish I knew enough about AI to set up and train a camera system. ;-)

[fedimakni]

Harrow
Thank you for your response. Did you try the capacitor approach before? could you share more details about the setup and the taken measurements?

[moffy]

If at all considering the camera approach, OpenCV, a large free computer vision toolkit might be of interest: https://opencv.org

[Harrow]

Harrow
Thank you for your response. Did you try the capacitor approach before? could you share more details about the setup and the taken measurements?

Sorry, no, I've not tried this. It was David's idea and I just thought it was a fun approach, but I've never tried it myself.

[David Hess]

For the capacitor plate approach I was thinking of the capacitor plates placed outside of the pipe that the fish is traveling through, with another pipe around them sealing them off from the water.  I was not thinking of allowing the plates to contact water.