Infrared 101

Light, Vision, and the EM Spectrum

In this universe of ours there is a ubiquitous form of energy we've given the rather scary name electromagnetic (EM) radiation. It forms a spectrum, within which a small slice can be detected by human eyes. This is "visible light", which spans the colors of the rainbow from red to violet. Beyond the violet end of the spectrum is ultraviolet (which humans can't see but other animals can), then X-rays, and Gamma-rays. In the other direction beyond red is infrared, then microwaves, radio waves, etc.

So where does light come from? For most of natural history, almost all visible light on earth came from the sun. This includes moonlight, which is just a reflection of sunlight. Exceptions include starlight, fire, lightning, molten lava, and an occasional oddball organism that figured out how to glow; and lately humans have added lightbulbs and televisions and the like to the mix. Most of these exceptions are chemical reactions, electrically-induced excitations, and other effects from throwing matter and energy at each other in interesting ways.

If we exclude such "active" interactions, where else do EM waves come from? It turns out that all matter emits EM waves all the time, a phenomenon called "black-body radiation". Of particular interest is that every "body" of a given temperature will emit EM energy around a given wavelength. In other words, everything glows, and the "color" of its glow is related to its temperature. This relationship is given by Planck's Law (actually Wien's Law, but Planck is famous and nobody's heard of Wien):

The X-axis is in logarithmic scale while the Y is linear, because the gentle curve is prettier to look at than a straight line or a sharp curve.

Notice, first, that for something to glow in the visible light spectrum, it needs to be around 4000-7000 Kelvin. This happens to be where you'll find the temperature of the sun, whose black-body spectrum peaks at yellow, is quite strong through ultraviolet, and drops off in the infrared. It's no coincidence that living things see in this range: it's what's most plentiful on this planet!

(You may also recall the astronomical factoid stating that hotter stars are blue while cooler stars are red. The graph above illustrates this nicely.)

Now, let's see where the spectrum lies for "room temperature" black bodies (300 Kelvin): it's way down in the long-wavelength infrared (LWIR). These are the wavelengths at which most things around us glow, and an infrared cameras at these wavelengths will let you "see" these temperatures.

NIR and Far

The infrared spectrum is quite broad, and you'll find the term "infrared" associated with technology in very different contexts, which might lead to confusion. The two ends of the spectrum that are most commonly encountered are Near Infrared (NIR) and Long Wavelength Infrared (LWIR). NIR, which is just beyond visible red, is common and inexpensive; it is used in TV remote controls and night vision goggles. LWIR, on the other hand, is at the far end of the spectrum at which room temperature black bodies glow. Infrared photography at the LWIR end is called thermal imaging or thermography, and is what this site is all about.

Night vision equipment uses infrared in various ways, but these are in the NIR range, and do not "see" temperature. "Passive" night vision equipment uses NIR light to brighten the image, but still relies on illumination from the moon or other ambient light; i.e., it only enhances what light is already there, and will stop working when it gets too dark. "Active" night vision equipment shines a flashlight at the scene, only the light is in the invisible NIR range. What you can expect out of such equipment is comparable to what you can expect with a flashlight at night: a cone of visibility, darkness outside the circle, occasional glare from reflective surfaces, range dependent on the power of the flashlight. Both of these technologies let you see shape and texture, but not temperature. It won't pick out a suitably camouflaged target, for example (with the caveat that conventional visible-light camouflage may reflect NIR differently).
Image from "night vision" goggles. (Source: Wikimedia Commons)

Thermography in the LWIR range is a different beast altogether: it is seeing each object's natural "glow", which varies with temperature. This means it will work in pitch darkness, for example, and that it will see through camouflage for warm-blooded animals.

So which is better? It depends on the application.

  • LWIR equipment is more expensive. Much more. And it's not just the sensor; lenses have to be made from exotic materials because LWIR does not penetrate glass. NIR detectors, on the other hand, are much cheaper; in fact, digital camera sensors already detect NIR, and camera manufacturers have to explicitly add a filter to exclude the NIR. There is actually a community of photographers that hack digital cameras to remove this filter, to get photographs with a dreamy effect.
    Infrared photography. (Source: Wikimedia Commons.)
  • NIR images look more natural, and are easier for humans to interpret, since it is basically just augmenting what we're already used to seeing. LWIR is very different, and takes some getting used to. In fact, the manner in which we choose to render LWIR images (such as the prototypical "rainbow" palette) is very much an art, and will depend a lot on application. (I intend to blog about thermal palettes in the future.)
    A distant squirrel in a tree, scene rendered with four palettes from the ThermApp app.
  • LWIR will find warm objects against a cooler setting. This is the feature that makes this technology useful for birding and wildlife detection. But once detected, LWIR will not produce a "better" image of the target. You can use a LWIR camera to find an owl at dusk, but to appreciate its beauty you need a flashlight or some form of NIR night-vision equipment.

In Conclusion

When I use the term "infrared" on this site, I really mean LWIR, thermography, or thermal imaging. I debated whether to switch terminology to avoid ambiguity, but those alternatives are both cumbersome and unfamiliar, so I decided to just stick to "infrared".

No comments:

Post a Comment