|
Forums36
Topics40,960
Posts557,936
Members18,497
| |
Most Online3,612
Jan 10th, 2023
|
|
|
15 members (Lumberman1985, gautprod, Boondoggle, Willowwood, Jward87, Sunil, Justin W, Jadog, Augie, Theo Gallus, phinfan, catscratch, Don Kennedy, FishinRod, Shorthose),
773
guests, and
264
robots. |
|
Key:
Admin,
Global Mod,
Mod
|
|
|
|
Joined: Jun 2008
Posts: 6,979 Likes: 14
Ambassador Lunker
|
Ambassador Lunker
Joined: Jun 2008
Posts: 6,979 Likes: 14 |
Love the idea of a light sensing trigger for the feeder. Light sensing, wind direction, and strength would be a hot selling add on.
Cheers Don.
Give me one with a rain gauge. My fish do not eat feed well when it's raining, so a feeder with a means to detect precipitation and suspend feeding operations until the shower passes might help me out. I know absolutely nothing about facial recognition systems, but my feeble guess is that such systems map the face, and compare it against an image stored in the system. Features might be reduced to measurements, which could easily be compared against the "stock" photo already on file. A certain percentage matches, 90-100% ??, and it says ok. Lesser percentage matched, no go. Building on that, let's have that stock photo be of the pond's surface on a calm day, and compare that image to one taken after an initial, test burst of feed hits the water. Analyze the difference caused by the feeding efforts of the fish. In this case however, if the new surface matches the stock photo, or within a predetermined percentage, then very little feeding is occurring and the feeder shuts down without dispensing any more food. But if the measurements and the math indicate a greater percentage of change AFTER the test burst hits the water, then open up and feed away. The problems and issues to overcome would be numerous. Windy days causing surface disturbances, leaves and debris falling on the surface, etc. And I'm sure JKB is right....it wouldn't be cheap. Still, fun to envision.
"Forget pounds and ounces, I'm figuring displacement!"
If we accept that: MBG(+)FGSF(=)HBG(F1)
And we surmise that: BG(>)HBG(F1) while GSF(<)HBG(F1)
Would it hold true that: HBG(F1)(+)AM500(x)q.d.(=)1.5lbGRWT?
PB answer: It depends.
|
|
|
|
|
Joined: Dec 2009
Posts: 6,692
Hall of Fame 2015 Lunker
|
|
Hall of Fame 2015 Lunker
Joined: Dec 2009
Posts: 6,692 |
Money would be the obstacle imposing the difficulty. I put 30 little HBG from Stony Creek in my Mom's former Koi pond. They have been hiding within the rocks of the waterfall since Saturday, only poking out once in a while at night. Tossed in some Zeigler 1/8" floating tonight and got a pretty good response for the first time.  Lack of funding always seems to be my major stumbling block. The HBG should be an interesting experiment. I'm torn between advising you to add a single LMB once you detect signs of spawning, or cutting back on the feed and letting them consume their own offspring. Of course with 30 fish you may not have over 2-3 females, so you could try sight fishing them out, once they grow large enough to cause a potential problem. It's just a teeny pond 
|
|
|
|
|
Joined: Dec 2009
Posts: 6,692
Hall of Fame 2015 Lunker
|
|
Hall of Fame 2015 Lunker
Joined: Dec 2009
Posts: 6,692 |
I know absolutely nothing about facial recognition systems, but my feeble guess is that such systems map the face, and compare it against an image stored in the system. Features might be reduced to measurements, which could easily be compared against the "stock" photo already on file. A certain percentage matches, 90-100% ??, and it says ok. Lesser percentage matched, no go. Maybe "fishal" recognition You could probably tie a number of these items into a Micro PLC that would not cost much more than a replacement timer. Only problem is you would need to learn how to program it. Some of this software can be a PITA and a tad clunky.
|
|
|
Moderated by Bill Cody, Bruce Condello, catmandoo, Chris Steelman, Dave Davidson1, esshup, ewest, FireIsHot, Omaha, Sunil, teehjaeh57 |
|
