Colorado
Offline
Full Member
Karma: 0
Posts: 140
|
 |
« on: September 20, 2012, 05:03:34 pm » |
I did some searching but found only posts in the old read-only forum. I thought there may be more current information and/or products, so posting the question here.
I'd like to connect a motion detector to an Arduino and have it point at my porch from behind a clear glass window. Someone approaches my door and the Arduino is triggered.
Can someone recommend a motion detector that works from behind glass, or recommend what I should be searching for?
Thanks, Jake
|
|
|
|
« Last Edit: September 27, 2012, 10:44:22 am by BlueJakester »
|
Logged
|
|
|
|
|
USA
Offline
God Member
Karma: 14
Posts: 644
|
 |
« Reply #1 on: September 20, 2012, 05:17:06 pm » |
Can someone recommend a motion detector that works from behind glass, or recommend what I should be searching for?
One thing to keep in mind is that common forms of glass aren't as transparent to infrared light as they are to visible light. Therefore IR based motion detectors, like PIR, might not be the best for this application. Other than that I don't have any specific advice.
|
|
|
|
|
Logged
|
|
|
|
|
the land of sun+snow
Offline
Edison Member
Karma: 81
Posts: 2135
|
 |
« Reply #2 on: September 20, 2012, 05:18:32 pm » |
Ultrasonic and PIR motion sensors will likely not work in your case, so you'll probably have to go to some kind of computer vision scheme. And probably will not be super easy to do.
I have read about people using simple CdS cells to detect motion [place in a tube for directionality, and program A/D converter to sense "rapid" changes], but these things are sensitive to all sorts of light-intensity changes, ie "false positives". Welcome to the world of signal detection.
|
|
|
|
|
Logged
|
|
|
|
|
Massachusetts, USA
Offline
Tesla Member
Karma: 97
Posts: 6382
|
 |
« Reply #3 on: September 20, 2012, 05:21:37 pm » |
If you have a location on the other side of the porch where you can put a reflector you could use a "laser tripwire" to detect objects crossing the laser's path. The laser and light detector would be inside the window. A small mirror would reflect the laser light back to the light detector.
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Sr. Member
Karma: 0
Posts: 427
"In this house, we obey the Laws of Thermodynamics" Homer J. Simpson
|
 |
« Reply #4 on: September 20, 2012, 05:23:33 pm » |
I agree with the comment about IR. I tried taking pictures using an infrared camera looking out through a window. What it picked up were reflections of heat sources inside the room on the glass but not what was outside.
Maybe this is a doge but if you can put small mirrors outside then you could reflect a light off them and detect the reflection. If the 'beam' was cut you could alarm.
|
|
|
|
|
Logged
|
|
|
|
|
the land of sun+snow
Offline
Edison Member
Karma: 81
Posts: 2135
|
 |
« Reply #5 on: September 20, 2012, 05:37:48 pm » |
Oops, I missed the point about the porch.You probably wouldn't need something as complex as a laser, but just a blinking "infrared" led, a mirror, and a pickup. The blinking would be used as a synchronous signal, and that would effectively eliminate stray background pickup.
Basically, 40-khz stimulation and TV IR-remote detector, although those are probably much more sensitive than you would need.
IR Led will help eliminate regular light problems. Note - normal IR Leds and sensors operate in near-IR range, about 880-940 nm, while PIR operates in far-IR range, about 10000 nm [IIRC], so you'd actually be dealing with light and not heat.
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Sr. Member
Karma: 0
Posts: 427
"In this house, we obey the Laws of Thermodynamics" Homer J. Simpson
|
 |
« Reply #6 on: September 20, 2012, 05:51:46 pm » |
johnwasser: We seem to have come up with the same idea, I did not see your post before sending mine.
BlueJakester: If you are going to use an IR LED check if you really can detect it though the glass. If you can then blinking sounds like a good idea, but make sure it is fast enough that a leg swinging through the beam will be detected.
|
|
|
|
|
Logged
|
|
|
|
|
Colorado
Offline
Full Member
Karma: 0
Posts: 140
|
 |
« Reply #7 on: September 20, 2012, 07:12:24 pm » |
Thanks everyone for the excellent suggestions. I like the laser - mirror - photocell idea. Sparkfun has some lasers like this one and I could reflect the beam into photocell attached to an Arduino. That would be a fun project to put together and learn from. If I do it I'll let you know how it goes.
|
|
|
|
|
Logged
|
|
|
|
|
California
Offline
God Member
Karma: 13
Posts: 631
|
 |
« Reply #8 on: September 21, 2012, 12:41:50 am » |
|
|
|
|
|
Logged
|
|
|
|
|
Offline
Jr. Member
Karma: 0
Posts: 57
|
 |
« Reply #9 on: September 21, 2012, 03:49:11 am » |
I used a PIR sensor module for the motion detection for a security light and fixed everything in a small transparent plastic box (PIR sensor module also inside the box). It is working fine without any issues. Whenever there is a motion it switches on the light. So I believe it works for your case as well with out any issues from behind the clear glass. Just give a try!!
|
|
|
|
|
Logged
|
|
|
|
|
Toledo, OH
Offline
Sr. Member
Karma: 16
Posts: 409
|
 |
« Reply #10 on: September 21, 2012, 10:00:23 am » |
Now that DOES seem interesting! Motion detection through walls! Love it, want! Tim
|
|
|
|
|
Logged
|
|
|
|
|
USA
Offline
God Member
Karma: 14
Posts: 644
|
 |
« Reply #11 on: September 21, 2012, 10:15:18 am » |
I used a PIR sensor module for the motion detection for a security light and fixed everything in a small transparent plastic box (PIR sensor module also inside the box). It is working fine without any issues. Whenever there is a motion it switches on the light. So I believe it works for your case as well with out any issues from behind the clear glass. Just give a try!!
Plastic an glass are very different categories of materials, and even different types within those categories can have very different optical or other physical properties. Furthermore, just because something is transparent to one set of electro-magnetic (EM) wavelengths doesn't mean it will be for another, even if the two sets are close to each other on the EM spectrum. Based on my own experience with IR sensors, in the mid-to-long infrared wavelengths (which is what PIR uses) window glass doesn't act like much of a window. There are of course types of glass that would be transparent to these wavelengths, but most aren't the type you'd get at your local hardware, art supply, or hobby store. Instead, you'd have to start with organizations that sell laboratory grade glassware and/or specialty optics.
|
|
|
|
|
Logged
|
|
|
|
|
Colorado
Offline
Full Member
Karma: 0
Posts: 140
|
 |
« Reply #12 on: September 27, 2012, 10:43:59 am » |
UPDATE: I received the parts I needed and finally had the time to setup and test in my shop. I'm directing this laser beam across the room into a mirror where it reflects back into this photocell. I placed a large diameter piece of heat shrink, about 1" long, around the photocell like a sleeve to cut down ambient light. At this time there is no glass window between the beam and mirror. When the beam is interrupted an LED illuminates. The reaction time of the photocell is plenty fast enough. The main loop() in my code is running without any delay. I can swing my hand through the beam in a karate chop motion and reaction is fast enough to trip and light the LED. I can flick my finger (like you were flicking a piece of sand off a table) through the beam and it triggers the event. I do have one concern so far though. Trying to line up that dot of laser beam in the mirror and direct it back to the photocell is tedious. The slightest touch or bump of either mirror or Arduino and the beam is no longer hitting the photocell. Any ideas on how to make that part easier? I also am going to need some sort of a small mirror on a swivel joint. The mirror I have now is about 2"x3" and no swivel. Something like a motorcycle rear-view mirror would work, but that would be way too big and noticeable on my porch. Any ideas? Thanks, Jake
|
|
|
|
« Last Edit: September 27, 2012, 11:05:04 am by BlueJakester »
|
Logged
|
|
|
|
|
Massachusetts, USA
Offline
Tesla Member
Karma: 97
Posts: 6382
|
 |
« Reply #13 on: September 27, 2012, 11:09:00 am » |
If you get an analog input from the photocell you can use that to generate a tone which will give you audio feedback for the aiming of the mirror.
|
|
|
|
|
Logged
|
|
|
|
|
Colorado
Offline
Full Member
Karma: 0
Posts: 140
|
 |
« Reply #14 on: September 27, 2012, 11:15:19 am » |
Thanks John. I reversed the event ... when the beam of light hits the photocell the LED lights. Now I can determine from across the room when the beam is lined up. Break the beam, LED goes off. If I put this project into actual use, I will put a timer on the LED to go out after a few minutes of making alignment.
It's still difficult and tedious to make that alignment. I suppose that's just the way it is going to be.
|
|
|
|
« Last Edit: September 27, 2012, 01:19:08 pm by BlueJakester »
|
Logged
|
|
|
|
|
|