Yes you have quite a few interesting points. I intended for this to be a basic project so I do not want to purchase a binch of accessories and such so perhaps this isn't the best project to start with.PiGraham wrote:That's an interesting and difficult challenge.
How big an area would you have to cover to detect eggs? Like 10 by 10 inches
Will the eggs be present when the hens are away? Yes
Will the eggs be on soft materials (straw?) Yes, but that is not necessary
Will the eggs generally be visible from above? Yep
Some quick thoughts:
Eggs contain water. Maybe the electrical properties of the water in the egg could be detected. 2.4GHz is absorbed by water. Perhaps a 2.4GHz radio signal would be altered enough by an egg to be detectable.
An egg weighs something, and much less than a hen. If you could weigh a laying platform perhaps you could detect when the weight goes up by a hen then down by a hen- an egg.
Eggs have hard shells. Perhaps the shell will reflect ultrasonic pulses more strongly than the nesting material they are in.
White eggs are quite bright and may show up well against straw, but brown eggs would probably be hard to see. Maybe egg shells fluoresce in ultraviolet light.
Freshly laid eggs will be warm. Maybe that warmth can be detected.
It's not going to be easy!
White paper fluoresces in UV (maybe brown does too). You could probably make a good mail detector with a few UV LEDs and photodiodes.Knightlorish wrote:I'm keen to do the same for my mailbox which is some distance away from the house and I'd thought of a change in image from the empty mailbox: easier than detecting an egg though - although a clever chap has achieved facial recognition - as included on a main page news item not so long ago.
Not always is it a disaster, just have to put in a little more effort to get things working for the creatures.aTao wrote:Technology meets animals = recipe for disaster.
I work on a farm, cows, pigs, chickens, sheep and turkeys. It is bewildering the amount of effort they all must put in to finding new ways to be contrary, sheep in particular seem to have a penchant for finding new ways to die. Chickens are also quite stubbornly unpredictable.
Having said that I would consider either sound (you try laying an egg the size of a beach ball (relative size) without making a racket) or infra-red image. A chicken will be warm but large, a freshly laid egg will be a sharp small(er) image that retains heat for considerably longer than the straw.
BTW, shouldnt chickens roost on a perch rather than in a nesting box?
That may not be a problem. Obviously the weight will flutuate. WHat you look for a drop where the hen leaves the box but the weight only drops by (hen-egg). If the hen is likely to leave the box and take an egg weight of straw and poo with it all at once the egg would not be detected, but if smaller amounts are kicked out over time it may still work. You would be measuring the hen weight every time it get on or off the box.aTao wrote:Trouble with weighing the box is chicken poo and straw gets kicked out.
That might work, I know such technology is used for material discrimination, but would the egg be masked by variations in moisture in the straw or variations in the distance to the egg?stijn.ghesquiere wrote:I've been thinking about making a eggs sensor myself.
I will use a capacity sensor that made to detect soil humidity.
Basically it consists of two small metal plates side to side (laying flat on surface) without making contact. You charge one with 5v or 3.3, whatever, the other is ground. Next you disconnect the loading and measure how low they keep charge. I'm not sure if a raspberry can do this fast enough, an arduino or a PIC processor can do so. You need to measure the very short times it takes to plates to unload.
The principle: you make a capacitor with this and you use the dielectric properties of water to 'store' a electric load. I'm sure you can even measure the size of the egg as larger eggs hold more water.
The actual implementation I'm heading for is to use a raster of metal plates to create what is actually a touch plate that can sense 'x-y' positions. I will use a PIC16f88 to do the actual timings and a serial connection to the raspberry pi to read the data and do things as calibration, position etc. in python.
No, not a "no", more of a busy weekend on my part and maybe the dreaded analog electronics for other people....stevetheipad wrote:Is that a no?
You can use USB for sound-in.stevetheipad wrote:I'm not experienced in electronics at all so I don't know much about that. I was hoping for more of a simple plug into the USB for cheap and use it in a python script.
Motion detection isn't very effective at detecting things that change slowly and don't move.Twinkletoes wrote:Could you feed the chickens with a mix of seeds and RFIDs, and then you'll get eggs with RFIDs built in. Unfortunately you won't be able to distinguish them from chickens...
Seriously though... Methinks an infrared camera would work best, with a motion detection algorithm to find hotspots that cool slowly and don't move.