sensor : detect human body in water

Thread Starter

Swashia

Joined Nov 24, 2012
4
Just like title said, I would like to build a device that detect human body floating on the surface of water, preferably with detection range of 50 meter. The first thing I come to mind is infrared sensor. Is it possible to build it cheaply? If so, what sensor chip do you recommend?

I found many infrared sensors chip online but they are mostly for motion detection
 
Last edited:

kubeek

Joined Sep 20, 2005
5,793
My guess is that an IR camera with some detection software might be able to do this.
Anything else like the IR sensor chips that you mentioned can only work if you know where the body is and have access to the whole water surface, and capacitive sensors won´t work at such distance.

Also, do you need to tell between a dolphin and a human, or just any large floating object will do?
 

spinnaker

Joined Oct 29, 2009
7,830
Does it have to be a human body? Or just something entering the water? A simple mercury switch on a float could detect something falling into the water.
 

MrChips

Joined Oct 2, 2009
30,618
Is the human body dead or alive?
Is the body of water a bathtub, swimming pool, reservoir, pond, lake, river, ocean?
 

strantor

Joined Oct 3, 2010
6,781
I'm assuming you're talking about some sort of detector for at-sea rescue. If that's the case, then this is not nearly as easy as you might think. The government and private sector are working on this, throwing tons of money at it, and still don't have anything very good. Millions of dollars spent on R&D, and they might be able to find you if they already have a good idea where you are, it's daytime, and the sea is relatively calm. Night time, in a storm, blown off course, not so much. Best bet is to keep a strobe light on you at all times if going to sea.
 
MrChips is correct in wanting some more parameters, namely is the body alive (producing heat), dead for a short time (heat somewhere above ambient), or dead long enough that body temp is same as ambient? Is the person wearing floatation (density also changes a lot as gas is produced in gut of a corpse), is there a water temp range you can limit this to, what kind of body of water (tub, pond, at-sea [echoing MrChips again])?
 

wayneh

Joined Sep 9, 2010
17,495
...I would like to build a device that detect human body floating on the surface of water, preferably with detection range of 50 meter.
It's called a helicopter. ;)

My point is, you need to specify 50m of what, horizontal distance? thru the water itself, from below? from above? calm water? waves and rapids? murky river water?

Maybe you need a shark
 

Thread Starter

Swashia

Joined Nov 24, 2012
4
Floating body implied that sensor have to detect any human being within 50 meters radius horizontally.

The human have to be alive so we can ignore dead body.

This is more like a prototype device built by the armatures so wide range detection is not necessary. We
are more like to test it on a small lake.

By the way 50 meters is fairly small distance, which is roughly the same length of Olympic swimming pool.
 
Last edited:

spinnaker

Joined Oct 29, 2009
7,830
I really can't see how this is possible without spending loads of money on the project. And still it light not come to anything.

It would be a huge job detecting an object near the surface at 50 meters. Sonar could be used but I could see where you would get a lot of distortion so close to the surface. But that is only a guess.
 

kubeek

Joined Sep 20, 2005
5,793
spinnaker, the sonar is a good idea, but I guess it would have to be deep (like 15-20m) underwater to get a decent "view" at the 50m radius.
Infrared camera on a pole would need less height, but still a lot, I think..
 

strantor

Joined Oct 3, 2010
6,781
I really can't see how this is possible without spending loads of money on the project. And still it light not come to anything.

It would be a huge job detecting an object near the surface at 50 meters. Sonar could be used but I could see where you would get a lot of distortion so close to the surface. But that is only a guess.
Yes, sonar isn't very effective and neither is radar. Thermal imaging might work for a few minutes on a fresh body, but water quickly turns a person cold. This is why it's such a challenge even for those with billions of dollars to throw at it.

It's such a challenge even, that I consider deploying sharks as a real viable solution. We've been doing it on land with bloodhounds for a very long time, why not in water? Sharks are the bloodhounds of the sea. There is still no technology that can beat a bloodhound's nose and smarts; that's why we still use them to sniff drugs, bombs, and missing persons. Why not send in the sharks? put muzzles on them, and a tracking device. could work.
 

wayneh

Joined Sep 9, 2010
17,495
I was halfway serious about sharks, for the reasons you mentioned. But I think the OP may be talking about fresh water. Less shark (or dolphin) options.

I could see sonar working, as far as detecting an object in the water. Forget the floating part - it could find the bulk of the body that's under the surface. But a log would look just like a body.

Infrared through the water would have a hard time similar to visible light. I doubt you could see a body at 50m with a mask on, unless the water is pristine. IR wouldn't do much better, I think. Tough problem.

Since we're talking about a live body, facial recognition comes to mind. Scan the surface for something that looks like a face.
 

strantor

Joined Oct 3, 2010
6,781
I could see sonar working, as far as detecting an object in the water. Forget the floating part - it could find the bulk of the body that's under the surface. But a log would look just like a body.
It was my job to track ships & other aquatic targets while serving on the submarine. I know a little about SONAR, or at least the type we had. The type we had (that .gov paid big bucks for) was pretty worthless for tracking anything that floated on or near the surface, and so was RADAR. RADAR likes to see things that stick up out of the water, like ships and sailboats. Floating logs, trash, and dead bodies, not so much. Maybe a purpose-built 'other type' of sonar could be tailored to see objects near the surface, but I doubt it. As I understand it, anything other than still-as-glass water is sure to throw it off, as SONAR uses sound waves, and ocean waves are not silent, and they move.

Infrared through the water would have a hard time similar to visible light. I doubt you could see a body at 50m with a mask on, unless the water is pristine. IR wouldn't do much better, I think. Tough problem.
Infared is a little further out of my realm of understanding, but I don't believe it would work through water; I believe it would only the temperature of the water that is directly in front of the lense. Could be wrong about that, but have you ever seen what happens when you try to look through a window with an IR camera? What's clear to us isn't clear to the IR camera.
Since we're talking about a live body, facial recognition comes to mind. Scan the surface for something that looks like a face.
That sounds a little star-trekkish, but it's the only thing I see in the ring contending with sharks with friggin LASER beams attached to their heads.
 

wayneh

Joined Sep 9, 2010
17,495
Maybe the key is to attach whatever sensor to a tethered balloon and get it up 20m or more, pointing it down at the water surface. Ultrasonic echo-location or IR might have a chance. It would be far easier if you could use a computer or a pair of eyeballs to help with the pattern recognition.
 

tshuck

Joined Oct 18, 2012
3,534
A LiDAR with knowledge of the pre-existing surrounding could do it. Although, you'd need to set it properly. You could set the system to only monitor points of interest(assumed humans) when they move above a certain rate(Δ0.05m/sec for example) and anything that moves slower is incorporated into the assumed surroundings...

There are, of course, some drawbacks to this, but it seems the most reliable to me...

Don't forget, sensor fusion makes a system more reliable.... so, LiDAR with mutated, fresh-water makos with lasers on their heads;)
 

strantor

Joined Oct 3, 2010
6,781
Maybe the key is to attach whatever sensor to a tethered balloon and get it up 20m or more, pointing it down at the water surface. Ultrasonic echo-location or IR might have a chance. It would be far easier if you could use a computer or a pair of eyeballs to help with the pattern recognition.
Now that sounds plausible. Balloons are cheaper to deploy than helicopters that need pilots and fuel. Maybe if they could be made to hold/change positions in a storm, a great many of them could be deployed to cover large expanses of water. Or maybe small electric helicopter drones could be mass produced.
 

gerty

Joined Aug 30, 2007
1,305
I volunteer with our county Search and Rescue and this is something nobody has figured out yet. We have an underwater infrared camera, visibilty depends on turbidity of water, sometimes 20' usually less. Sonar gives all kinds of false alarms, usually partially submerged logs. Dogs that we've tried were cadaver dogs, body must've been in the water several days. Cadaver dogs sniff out the excaping body gasses, a lot depends on thier last meal and ambient temp.
What the person is wearing has some effect on the bouyancy of the body also. Heavy winter clothing thats absorbent will weigh the body down.
 
Top