Where do I find transparent coatings that show up in near-infrared photography?

April 16, 2009 11:46 AM

I need a non-toxic coating that is transparent under visible light but appears much darker/lighter in near-infrared photography.

My dear chemist, physicists, and material scientists, I am working on a project where I want to coat uniform surfaces with a substance that appears transparent or semitransparent in normal human vision but shows up really well under near-infrared photography. My infrared camera is a USB web cam with the IR filter replaced with some flavor of IR-Pass filter. I should be able to make my own paint if you can point me in the direction of a pigment that will have the desired results. Bonus points for substances that are non-toxic.

I do computer vision work and I want to detect image features on substances that would otherwise look fairly featureless (like a wall, or a tabletop).

posted by kscottz to science & nature (23 comments total)

What about something like UV-reactive fluid, like what they use for blacklight handstamps at clubs?
posted by sadiehawkinstein at 11:58 AM on April 16

Yeah, the other (UV) end of the photo spectrum might be easier in terms of materials.

Hell, you can even use spray on UV sunblock.
posted by ztdavis at 12:07 PM on April 16

Isn't the whole point of IR photography that it sees temperature differences? I don't see how a transparent coating is going to significantly affect the temperature of the surfaces you apply it to.
posted by jon1270 at 12:09 PM on April 16

What about something like UV-reactive fluid, like what they use for blacklight handstamps at clubs?

I already tried that and it doesn't work. Light coatings of metallic spray paints kinda work but only in the sense that they are reflective in the IR spectrum. I haven't found a marker or coating that works well yet.
posted by kscottz at 12:10 PM on April 16

Isn't detergent UV-reactive? Or was that just LOST...
posted by miratime at 12:11 PM on April 16

You might have some luck contacting the folks behind this website, although I don't know if it's exactly what you are looking for.
posted by pemberkins at 12:12 PM on April 16

Yeah, the other (UV) end of the photo spectrum might be easier in terms of materials

I agree UV is much easier to work with. The problem with UV is most cheap CCDs don't pick it up that well (at least that is my understanding). To use UV paint I would probably need a specialized camera, or to use a UV light source so I can see the UV paint in the visible spectrum, which would defeat the purpose.
posted by kscottz at 12:15 PM on April 16

This looks like it might be a promising product - I don't know if you could actually procure it from these folks or not, but I will post it in the hopes that it will inspire someone to find a more convenient source of something along these linge.
posted by pemberkins at 12:20 PM on April 16

And by "linge", I mean "lines." Sigh.
posted by pemberkins at 12:20 PM on April 16

Glass is opaque in the infrared, but transparent in the "optical" part of the spectrum. Glass coating is out, but I wonder if small glass squares as markers might suffice for the time being.

Underdrawing with carbon black isn't quite what you want. I gather you would like to have something you could spray on or otherwise easily apply. With underdrawing, you'd have to apply the carbon black, then apply white paint on top of it.

This might be helpful.

I wonder if clear primer might not do the job.

Damn, this is a good question.
posted by adipocere at 12:25 PM on April 16

@jon1270 Actually kscottz is aiming for something different. He or she is working with machine vision, using a camera that can see only really bright infrared, the kind that is emitted by the LED in your television remote, for example.

In order to measure temperature with infrared, like you're talking about, you'd need much more sensitive equipment.

A picture is worth a thousand words: http://www.nuigroup.com/uploads/cross.jpg

...Of course that picture isn't specifically what kscottz is going for. At least it's in the ball park.

I believe that kscottz is attempting to create fiducials that are invisible to humans but visible to his camera.

@kscottz, have you considered layering your paint (or other material) backwards? Like, put infrared reflective tape down in whatever pattern, then cover it with a visible-opaque, IR-pass material. (Like the black plastic on the front of a remote control.) If/when you do find something, please mention it here so I'll know! :)
posted by DavidLoyall at 12:25 PM on April 16

jon1270: "Isn't the whole point of IR photography that it sees temperature differences? I don't see how a transparent coating is going to significantly affect the temperature of the surfaces you apply it to."

I think what we're talking about here is near infrared - the kind that makes for stunning photography because the absorption / reflection of many materials (plants, water, stones) is quite different for NIR than for the human-visible spectrum. Google infrared photography for more info, or check these sites for examples: short introduction on the topic; some example images (sorry, German text); some more; more in-depth FAQ with some nice side-by-side comparisons of regular and NIR photography; another picture gallery.
Errr, sorry, I was kinda interested in that stuff for quite some time.

As to the original question: I'd either ask around on any of the IR photography sites if anyone had noticed any particularly striking paint that might fit your needs; could you try using a chlorophyll-based pigment (might not be exceptionally durable, but should show the Wood effect nicely)? Other than that a quick google led to this site which seems to be a research group specializing in coatings that do just that.
posted by PontifexPrimus at 12:38 PM on April 16

DavidLoyall,

You hit the nail on the head, but I can back layer fiducials on the surface. I want to take an existing object like a white wall or some house hold item, cover it with some "magic clear spray" and the automatically get fiducials in the IR. The best solution would be water cheap, non-toxic, water soluble, and be something you could get without special ordering it.

I am hoping that a chemist can tell me about some magic polymer or other chemical that fits the bill.
posted by kscottz at 12:42 PM on April 16

Some visible light blacks are white or gray in IR and some blacks are black in both visible light and IR. You could paint your objects with black paint that is white in IR and then add markings with paint that is black in visible light and IR. If you tested a bunch of pigments you might find some good ones to use.

Check this diagram out. It shows a Canon "Hybrid Infrared-Cut Low-Pass Filters" they mention "infrared absorption glass" if you could get your hands on some of that fancy glass you could powder it and put it in a clear glue and paint it on...
posted by gregr at 12:43 PM on April 16

gregr - I might see if these guys can give me the by product of their lens grinding. I wonder if this stuff is anisotropic? Does the effect diminish if you grind the glass into little bits that face all directions.
posted by kscottz at 12:59 PM on April 16

kscottz - I'm not sure that glass will work; it's absorption range is listed as 1700 to 2500nm. Your cameras probably aren't very sensitive or sensitive at all in that range.
posted by gregr at 1:38 PM on April 16

Wouldn't whatever the IR filter that was originally in front of your CCD was made of fit the bill?
posted by floam at 2:00 PM on April 16

What is the IR wavelength range you can sense? Fused silica absorbs at 1126, 809, 452, and 200 cm−1

Long shot: This might not fit into your requirements, but what about glass micro-spheres added into the transparent coating? They make great retro-reflective surfaces. They have the strongest return at a 90 degree angle of incidence, so if your IR source was co-located with the camera lens, these would shine like crazy. Is IR reflectance okay, instead of absorption (bright fiducials instead of dark)? The obvious problem is that the spheres will do the same bounce back in the visible range for all other viewing that take a place near visible light sources in the room. Couldn't tell if that is bad for your application.
posted by buzzv at 3:39 PM on April 16

Grinding up that Anchor glass seems like a good way to go. Edmund Optics apparently sells it and has an online price list:

The IR spectrum looks flat (image from Edmund product page).
posted by buzzv at 3:47 PM on April 16

If a suitable dye cannot be found, you might be able to figure something out with antireflective coatings. They're wavelength specific - often centred on the green for the optical but one could get NIR-tuned coatings.
Cost might not be practical however.
posted by edd at 5:32 AM on April 17

Use the stuff they sell to put on license plates to confuse speed sensing/imaging cameras.
posted by gjc at 6:01 AM on April 17

Neat question and I wish I could help. I'm sure you've already checked out the wikipedia page so I'll just note two things on that that you might have missed: plants are almost always blazing white in NIR photos so you might want to ask a botanist/biochemist if they know of anything. Also on the wikipedia page, the 5 dollar bill photo clearly shows someone out there knows how to do something like what you want. I don't think the U.S. Mint will help you though.
posted by chairface at 10:28 PM on April 17

... or you could click first link when you search for "IR ink"...
posted by chairface at 10:31 PM on April 17