3004Views15Replies

Author Options:

Human Detector Answered

I'm trying to come up with a small, cheap, simple circuit that will detect a human. I plan to use a grid of luxeon rebel OLEDs and whatever detector I come up with embedded in my ceiling to light up only when you are under them. I've looked into passive infrared, but it's fairly expensive for the components I could find. Any leads for circuits, ideas, or parts? I'd use parts from cheap security lights, but they tend to be around $10 each and I would probably need around12-18 ... kinda painful on the pocketbook. Thanks in advance for any help!

15 Replies

user
Toga_Dan (author)2016-07-19

quiz the potential human to see if they can recognize text which is distorted in a funhouse mirror way. Most dogs, cats and robots can't correctly read such lettering, and so, they don't deserve light.

Select as Best AnswerUndo Best Answer

user
KanjanaC (author)2016-07-19

Imagine the follow situation:

An auto-car driving down the road, then:

in front of the car there is a human being, on its left there is a Pig (weighting about 180 Kg), on its right there is another human being (a fat one - weighting just about 180 Kg).

We all know the car can't stop in time and one must be targeted. what will be?

you have to run some background test, heat detection(infrared - but only is not enough), size measurement, form detection (It needs in some way looks like a human being) and a big dose of Morality (machine learning)!

Select as Best AnswerUndo Best Answer

user
homba (author)2008-01-30

Kinda replying to all in one post here ... No pets or autonomous robots :) I've been thinking of ir beams for our outdoor sidewalks, but indoors I think that I would have to have a really tight grid or someone might get missed. Also, people may sit down or move tall furniture in the way. I've been thinking of some sort of IR, though. Why would I want to avoid PIR if I can find cheap enough components? $0.99 would probably fit the budget. I saw the goldmine-elec parts, too, but was wrried that I couldn't find a spec sheet for it. I'm very open to other ideas, though. I've been toying with the idea of antenna detectors like a theremin, but It seems you have to be pretty close to the wire for any detection.

Select as Best AnswerUndo Best Answer

user
gmoon (author)homba2008-01-30

At .99, you could probably reverse engineer the detector (based on other PIR schematics), even if you bricked a couple.

One point--you could reduce the number of sensors by widen the detection area of each one. Detection response could be based on whether you're triggering 1,2, 3 or 4 sensors simultaneously.

More of an interconnected mesh, than individual autonomous sensors. More logic complexity and wiring, but probably cheaper...

Select as Best AnswerUndo Best Answer

user
homba (author)gmoon2008-01-30
That's a good point regarding a mesh-type grid. I wanted to do analog only, but a pic would probably be easier. I used to be quite a competent programmer in a past life, but my circuit-designing skills are pretty weak.

I have a basic stamp I dev board That I can usually get to do what I want. I could probably use 4 detectors to control maybe 9 lights.

like this, perhaps?
  • = light
x = detector:

  • * *
x x
  • * *
x x
  • * *

Select as Best AnswerUndo Best Answer

user
homba (author)homba2008-01-30
grr ... one more try
o = light
x = detector
o     o     o   x     xo     o     o   x     xo     o     o

Select as Best AnswerUndo Best Answer

user
gmoon (author)homba2008-01-30

Yeah. I'd tend to use a micro. But I bet you could do it with discrete gate logic, too. The devil would be in the details ;-). Sure would make an interesting exercise in logic, though...

Select as Best AnswerUndo Best Answer

user
homba (author)gmoon2008-01-30

I really like the discrete logic idea. I plan to use this i'ble to drive the OLEDs, so I think it would mostly be a matter of testing out the PIR detector for voltage/resistance/whatever, designing the logic, and then figuring out all of the values needed and sourcing the parts.

My big hangup is figuring all the component values. I lost my electronics book in a move (I'm moving again in a month and hope to find it again). But even with the book, I tend to guess values. I can usually fudge something together for a small/cheap circuit, but not something complex like this. Any links to sources to better educate myself?

Select as Best AnswerUndo Best Answer

user
homba (author)homba2008-01-30
dangit! try this:
{{{x = detector}}}{{{*         *        *}}}{{{     x          x}}}{{{*         *        *}}}{{{     x          x}}}{{{*         *        *}}}

Select as Best AnswerUndo Best Answer

user
thejrb (author)2008-01-30

You would probably want something that would beep or flash or something when it senses heat at the body temp. No clue how you would do it but a start.

Select as Best AnswerUndo Best Answer

user
gmoon (author)2008-01-30

You don't want PIR, but I've seen the detectors and boards sold very cheaply, if you look around-- here's an example...

You probably already saw the PIR detectors for robotics 'ible posted here just days ago....

Select as Best AnswerUndo Best Answer

user
gmoon (author)gmoon2008-01-30

Elec Goldmine also has the detectors alone for 0.99 ....

Select as Best AnswerUndo Best Answer

user
schorhr (author)2008-01-30

A couple of IR-Beams across the hallway in shoulder-height might do it. If a beam is blocked, someone must be present. Dont use the sensors as coathangers though.

Select as Best AnswerUndo Best Answer

user
Patrik (author)2008-01-30

Do you have any pets - or a roomba? If not, that simplifies things significantly!

Select as Best AnswerUndo Best Answer

user
Goodhart (author)2008-01-29

To make the detection you have to detect what is unique about a human (as opposed to say, a bear, dog, deer, automobile, etc). There is size (but then we have bear and a few simians) warmth to separate out the non-living....what will be the method of determination / seperation?

Select as Best AnswerUndo Best Answer