Water/land detection/differentiation from a quadcopter

[lostminty]

Hello,

I want to have a quadcopter drone type set up which can tell how far from land it is. I would go for GPS but I want it to take into consideration the tide, and in general I'm only talking a distance of 50-200m (within shore fishing distance).

The ideal system would be a drone that can traverse along a beach front a set distance from the shore line. Some of the newer drones have sonar for use inside homes but I doubt that is very useful. I'm not that up on the play with what I could bounce off the surface that would show a difference between water and ground. Maybe microwave? I expect though that will have a lot of scatter.

My first instinct is image analysis. Perhaps using an IR camera? I've heard that I think CMOS sensors already detect IR and normally have a filter in place. I haven't used image analysing algorithms before. But I'm hoping once I have a drone I can film the shore line to see whether tweaking images produced will allow for a processing friendly derivation of the shoreline position.

 

[Baluncore]

The method will depend on the environment. What is your latitude? Is there snow on the ground or is it a tropical beach? Night or day? You might consider an FLIR camera to detect the temperature difference between the water and the beach. You would need to know your height and analyse the image to get a range.

If there is wind and waves there will be a microwave doppler flutter from the sea surface, but not from the beach, unfortunately there will also be bits of sea that have no doppler flutter.

A laser would back scatter from the beach and give you a range measurement, but it would reflect away from a flat water surface because it has a low angle of incidence. The laser would need to keep searching for the boundary. Breaking waves that were visible to back scatter light would probably be close to the beach, depending on sea state.

If you used a vertical microwave or laser you could measure the depth of the water by reflection from both the sea surface and sea floor. That method would permit following a water depth contour rather than a fixed offset from the beach. A chirp signal would show two time delay peaks in the processed spectrum. One would be the height ASL, the other the water depth.

 

[rollingstein]

What is the application? i.e. why does it need to know & how accurately? Is it the shortest distance to land?

 

[lostminty]

The idea is to suspend a fish finder below it, you can currently get castable fish finders. Then drag it across the beach head and get a rough idea of where the fish are. There will be a issues with drift of the buoy, it would be quite convoluted to get the drone to compensate for this with counter movement. I guess you could play with it using the load on the drone from the buoy. Either through strain gages or preexisting systems.

The FLIR camera was sort of what I was thinking. I believed the temperature difference between sea and land would be fairly distinct so that might yield a high enough contrast.

One will know the altitude, at least...some of the quadcopters do have altimeters.

I like where you're going with analysing the floor depth relative to the surface. I thought microwave was highly absorbed by water so I didn't pursue it far.

Ideally I would stick to off the shelf optics

 

[rollingstein]

The fish finder is Sonar, correct? Does it also get you a depth sounding?

 

[lostminty]

Good point. To which I cannot get an exact answer. They do not mention it. Many of the fishfinders pair via bluetooth. Some of them you have to manually set their depth so that it improves the quality of image.

 

[Baluncore]

A technology race usually signals the end of a resource. Once individuals or concentrated groups of animals can be located and tracked, an efficient hunting technology can finally extinguish the species.

Would it not be better in the long run to allow the fish population density to increase to the point where fishing with old technology yields acceptable economic results. The best way to find the optimum operating point is to avoid higher technology hunting practices now.

 

[rollingstein]

Good point. To which I cannot get an exact answer. They do not mention it. Many of the fishfinders pair via bluetooth. Some of them you have to manually set their depth so that it improves the quality of image.

Why not buy a dedicated sonar n also mount that on the bouy you will drag? Then you get authentic depth soundings very rapidly & can use those to follow a depth contour.

Seems a lot less work & more accurate than image processing. Besides you are already dragging an instrument. Might as well add another. Sonar sounds cost competitive versus an imaging system.