Testing IR Camera Blocking

[randy] from F.A.T. tested the theory that infrared LEDs can actually hide you from the prying eyes of surveillance cameras. We’ve previously covered camouflage, IR, and other suggestions for eluding the cameras, but haven’t taken to sewing stuff onto our clothes yet. [randy] lined his hoodie with high-intensity infrared LEDs, hoping to create a halo effect that would hide his head, and tested his results. Unfortunately, his efforts were unsuccessful. He tested many many different combinations and we’re confident in his conclusion that it would be very hard to make this work.

36 thoughts on “Testing IR Camera Blocking

  1. You know standard LED use 20Ma, but IR ones are meant for pulsing and can take 100 to 250Ma.
    In other words this guy needs to get a clue and get those things pulsing and(or) supply more amperages to them to get that stuff working because obviously using weak IR LED on a camera that has some IR filtering too isn’t going to work.

  2. This is stupid.
    He’s using the wrong kind of camera for testing.

    If you’d try it with a proper surveillance camera in nightmode it would work much better.

    Also, it would only work in very dark areas with highly IR sensitive CCTV cameras in nightmode.

  3. Thank you for posting this! I was doing tests myself and finding no results…. at least the IR LED is still good to make a light pen for the wiimote projects… right? hmm.
    It seems the easiest way to confuse security cameras is still the laser pointer in the lens.

  4. Because none of them have a clue as to how the cameras work.

    Go online and get 6-8 1 WATT IR led’s. THEN you will have enough of an IR light output to completely obscure your face on the cameras. the dinky toy led’s these guys are playing with will do nothing, you need a crapload of light output to overload the camera through it’s IR filter.

    Now at night when the cameras are in night mode and looking for IR light you do not need as much to overwhelm them.

  5. You could use lower power leds such as he is using if you defused them through something else such as the back lighting plane from a lcd display. to create a mass of ir instead of a single point.

  6. Right, there’s no evidence here. This guy is doing wrong tests. He should try with high power leds, pulsing high current and also buy WIDE angle (and not directive)leds !!!

  7. It seems like there was still a fair amount of ambient light during these photo’s. A key part to getting this to work is to make the camera’s exposure time too low to gather reflected light from the area of your face. Placing bright non-direction lights near your face will not help. I don’t believe this is an unfeasible task. If I were in a situation to need something like this, I’d go for IR spotlights like they used in the movie The Inside Man.

  8. anyone that says the camera is at fault, is seriously not thinking in the right direction. As a hacker, you do not have control over the camera used. with that said, he definitely has no idea how to find quality IR LED’s. the ones he’s using are inefficient at outputting the maximum ir spectrum.

  9. Ok, so I put 12 high intensity 5mm IR LEDs on some breadboard (had more LEDs but 12 seemed reasonable), got a few of my CCD cameras out, held the breadboard to my forehead and made this 1 minute video:

    Filmed in low light conditions (that’s a streetlight in the window).

    The first camera is a colour Sony 0.5 LUX CCD bullet camera, with it’s standard IR filter.
    The second camera is an identical bullet camera but I removed the IR filter from the CCD.
    The third camera iw a black/white 0.04 LUX camera just thrown in out of curiosity.

    As for hiding your face from cameras with IR filters? I’d hazard a guess that is’s possible but you’d probably need a significant amount of high intensity wide angle IR LEDs surrounding your face. I could have a stab at the ‘problem’ further but I’d rather spend time sticking the IR LEDs and a camera to my Sony LCD glasses – instant nightvision goggles!

  10. He didn’t use high brightness LEDs in pulse mode.

    If he used a luxeon IR LED the results would have been different.

    Also an IR laser out of a CD burner might work well too but this is dangerous since a near infrared laser powerful enough to kill a camera also can fry someone’s retinas.
    BTW cameras in theory should have a second weakness they do tend to be scanned like tvs the analog kind anyway so maybe transmitting an RF pulse that looks like a sync signal could mess it up.
    The digital kind maybe try and make the pulse looks like the signal encoding for USB or ethernet.
    802.11 wireless cams would be far easier to jam then any wireless cameras you wouldn’t even have to bother with LEDs just 3 watts worth of white noise on the 2.4GHz ISM band should do the job.
    Other frequencies of interest 2.45GHz and 900mhz for analog wireless cameras.
    I also wonder if near UV also could be used instead of IR though this would make many materials fluoresce.
    BTW this kinda reminds me of the optical jamming used in the tv show Phantom 2040.

  11. Anyone thought of pointing the LEDs at your face rather then at the camera? That way if you’re sideways to the camera the IR will still bounce off your face towards the camera – directed LEDs would be pointing off the same direction you’re looking and your profile could be visible to the camera. Harder to achieve though.

    And use high power LEDs, kids.

  12. lol damn… well this dude just tried what I’m sure many of us had planned on using in the event we wanted to obscure security cameras… maybe DIY IR laser pointer?

    you could use a camera without an IR filter to help with the aiming…

    better than a normal laser pointer, cuz no security personnel would know where the lights comin from…

    (I’m tired as hell… This all probably sounds retarded)

  13. I’ve thought about doing something like this but modulating the IR output with a wideband, psuedo-random sequence (which is super easy to generate with a few logic chips and a clock source). My thought is that this broadband optical noise might interfere with the camera’s scan rate, or other processing it does (similar to what ‘nitori’ described).

    Best case would be that the field of view of the camera would turn to ‘snow’, worst case, it would do nothing..

    ..and for everyone beating up on this guy for LED selection.. at least he tried something.. I give him props for making a reasonable effort in the first place…

  14. CCD-Arrays are a relatively high sensitivity in the IR range, but IR light generally degrades the image. Thus high quality industrial grade cameras are equipped with an IR filter directly placed before the sensor.

    The IR-LED approach works well with bright LEDs, but is naturally futile with cameras that filter their wavelength.

  15. Try using “near IR” LEDS. LED’s that are closer to the optical range than many that are used in remote controls. The 475-1112-ND (880nm) suggested by Charlie are closer to the Red spectrum so they may be at the edge of the IR blocking filter and work better. The cheap LED’s that I have seen on many sites are also diffused so they spread the light in more directions and have higher losses. Great for remotes but crap for this project. Of course, these will also be more expensive.

    I would also look at modulating the signal as well with an AVR. Something I am going to test.

    Fun project.

  16. I tested this theory over a year ago and was very successful… I used an infrared tree camera a.k.a. trail camera that hunters use, I super glued regular LED bulbs to me, mind you the bulbs had no power whatsoever to them, and were NOT INFRARED… and all you could see in the images from the trail camera was a huge blur of light… try it

  17. With most security cameras there are integrated IR lights in order to light the scene.

    I can’t say if it would work for this, but even a 1″ hat band made of retroflective tape, will blind a camera with its own flash.

    I used some which is black under normal light, but reflects the light hitting it directly at the source, since it doesn’t matter what kind of light hits it, nor the angle, it works even when you move your head.

    Facial recognition can be fooled in other ways, but unless you actually blind the cameras (400mW IR laser directly hitting the lens…or perhaps a bright enough IR pulse…or low tech, shoot it out with paint, paintball, wasp spray, etc.or destroy power, signal or recording/viewing systems,) such systems can collect a fair amount of information, which should also be obscured: gait, height, stride, arm length, balance disturbance, joint locations, et al. as these may be enough to identify you, and will, at least, narrow the suspect pool.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.