Poor man’s thermographic camera

thermographic camera

Using an IR thermometer, there are two ways to go about building a thermographic camera. The first uses a pan and tilt head. Scan lines are emulated, as a computer controls panning from left to right, taking a temperature sample from each step. Vertical resolution is accomplished by tilting. Another method uses a web cam attached to the thermometer. The thermometer’s laser pointer is captured with temperature annotations, as the computer records the field of view. We think the best outcome can be found with a combination of both methods. The video embedded below demonstrates the results. This would be a good addition to the Autonomous paintball sentry.

[Thanks Rolf]

Comments

  1. Haku says:

    Great stuff, I’ve been wondering how long until I see someone hack a handheld IR thermometer in this manner, mounting it on a couple of servos to scan an area & create a rudimentary thermal image.

    I’m still planning on making my own DIY thermal imager using a thermopile, servos & a PICAXE hooked into a computer. It may not produce the same quality images as a commercial thermal imager, but it’ll cost a fraction of the price.

    TPA81 8 point thermopile with i2c interface: http://www.totalrobots.com/tpa81-thermal-sensor-array-p-328.html

  2. eric says:

    Hey this is pretty cool, I was thinking about doing something similar with a single photo resistor or photodiode and use a pan/tilt to scan and area to take an image without a camera!

  3. Davandron says:

    ::Laughs::
    I always think its so cool when other people have the same idea that I had. But in this case, the person is way smarter than I; I hadn’t thought about the spatiotemperal approach to make random samples into a picture

  4. ex-parrot says:

    Is the sensor’s gaze susceptible to redirection by a mirror? If so a stationery unit could scan a whole area very quickly / easily.

  5. Tim says:

    ex-parrot: Yes if you use the right material. I seem to remember glass works.
    I think you couldn’t scan really fast though because of the integration time of the sensor.

    IR thermometers aren’t that great anyway. The emissivity of things varies so much it’s hard to get an accurate reading.

  6. ino says:

    Interting project.

    And for once, a nice background sound !
    So unusual in those kind of vids
    kuddo for that !

  7. David says:

    Many digital cameras are sensitive to IR. Point a remote control at your digital camera, can you see if you can see the IR LED pulsing? If so, put a camera IR filter over the lens and methinks you have an IR camera; right? Haven’t tried it myself; don’t have an IR filter. But they’re pretty cheap on ebaY.

  8. therian says:

    David infrared is not heat yet

  9. ex-parrot says:

    though hot things will give off infrared – point a digital camera at an electric stove element for instance.

  10. JB says:

    Nice hack. It does not have the resolution of a real IR imager (I have access to one at work) but should do fine for a home lab. I’d like to see more details on how to build it/software. As Steve said, it would be nice to add to a sentry gun (follow the heat and shoot until is no longer hot) :P

  11. Haku says:

    I think the most complex thing about the first version of this hack (IR thermometer on a pan/tilt setup) is the software, you can get IR thermometers with a USB connections so no need to start taking one apart to figure out how to tap into it’s circuitry.

    The thing that suprises me is how he only sampled a load of random points rather than scan a grid of points, unless that proved too difficult (unlikely) or too time consuming (likely).

  12. duesentrieb says:

    Althoug CCD cameras can be converted to see *near* infrared, the cannot be used to detect heat. Thermal radiation is so far away from the receptive spectrum of the CCDs that this would be impossible. The reason that some stove plates also light up in near IR is that they emit *some* radiation in that region, too.

  13. andar_b says:

    Yeah, you get them hot enough and they emit visible light with no trouble, of course they would pick up on a CCD.

    I had trouble understanding this for some time, because they’re used interchangibly, but its like dogs and poodles. All poodles are dogs, but not all dogs are poodles.

    Likewise, heat is infrared (below the wavelength of visible light) but not all infrared == heat.

  14. Younata says:

    Alternatively, you get a ccd.
    You put a filter which blocks all but IR.
    Then, using physics, you interpret the data you get so you get a nice looking picture.

  15. tom61 says:

    younata, you’d only be able to see a little cooler than red-hot with the IR range a CCD/CMOS sensor can see. The IR that consumer camera sensors can see is in the near-visible light range, ‘heat’ IR is farther away from visible light in the spectrum and cannot be seen.

  16. Gio says:

    Weren’t the old thermal imagers build up from a single detector and a system of rotating mirrors to scan a 2D image? Reflecting IR (8µm, mid IR, the region of thermal IR emission) is not the same as reflecting visible light, but I believe that material as common as aluminum should do the trick (for the mirrors of course). But maybe the sensor will react to slow to be useful.
    Does anyone know a manufacturer of sensors that are sensitive to 8µm wavelength? Most common IR sensors are only sensitive to 2µm tops.

  17. Gio says:

    Hmm, I was just thinking, maybe a PIR sensor from a cheap motion detector? But again, they might be slow. Anyone gave this a try yet?

  18. Haku says:

    From what I gathered of searching for ir thermal sensors, the ones used in motion detectors only detect heat changes, not the actual level of heat in celcius/fahrenheit.

    Which is a shame because I picked up a bunch of battery powered PIR sensors for £3 each, being battery powered they can be directly plugged into PICAXE chips and powered from the same 5v supply, hence my reason for getting them (that and the price is unbelievably cheap).

  19. Haku says:

    @Gio: the thermopile I linked to in the first post of this comments section works in the 2um-22um range. The price isn’t very nice but it’s the only commercially available sensor like that I could find which you can order online, similar ones I’ve found online are on manufacturers sites and don’t have online shopping capabilities.

  20. Thanks for the praise of my video. Actually, I’ve got tons more of such ideas, ranging from computer graphics and audio to Web-based stuff to electronics:
    http://www.j3l7h.de/ideas.html
    I’m definitively looking for international cooperation with students and with other educators/researchers. I’m traveling a lot to North America and to the UK so there should also be a chance to meet in person. — Joern

  21. Fred says:

    Great idea and execution!

    I am currently looking at something similar, but my infrared thermometer does not have a computer link. The ouput I get from the thermometer is a bunch of 1’s and 0’s. Any ideas on how to get some proper data from the thermometer onto the laptop or how to interpret the code?

  22. Antonio says:

    Can you use the ir in your tv remote to created an ir thermometer?

  23. ronald says:

    no thermal images like them thermal heat sencing night scopes

  24. nobodyimportant says:

    Antonia, an IR remote is an near infrared light source (just below visible red). You would need need a detector to see.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 91,398 other followers