I recently heard a homicide detective testify in a criminal trial that, if two individuals appear side by side in an infrared video recording (hereafter “IVR”) filmed at night, and one appears to be wearing a dark jacket and the other a light jacket, the clothing will not necessarily appear darker and lighter relative to each other, if seen in natural lighting. That is, the “darker” jacket might actually appear lighter than the other jacket, if seen in daylight.
That surprised me. Of course color doesn’t show in IVR, but I would have thought that relative darkness and lightness would remain the same under natural lighting.
Anyone with experience with IVR who’d like to weigh in?
Hotter things show up lighter in infrared imaging, cooler things show up darker. The color of clothing doesn’t matter, a hot black jacket shows up lighter than a cooler white jacket. Clothing is often going to insulate body heat which is going to usually be hotter than any clothing but it will depend on what other sources of infrared are around.
More important will be the thickness and type of material the jacket is made from. A thinner jacket will let out more body heat, for example. This effects how it shows up in an IVR. The relative temperature and IR reflectivity of the clothing is more important than the visible color of it.
Hold on, Nellie!
By “Infrared,” do you mean “FLIR”? Because that is a whole lot different than “night vision.”
FLIR creates a false-color image of the temperatures of objects. The temperature of various objects in the scene appear in a rainbow of colors:
Night Vision is a scene illuminated with near-IR LEDs, and renders the scene is shades of grey (usually). The reflectivity of objects to near-IR determines how “bright” they appear in the image.
Some things (including some clothing) are almost transparent to near-IR of the type used for night-time video recording. How an object appears on that type of camera will depend on whether it absorbs, reflects, or is transparent to that particular band of wavelengths - it’s quite possible for two materials to look very similar under normal visible white light, and look very different from each other under near-IR.
But would a dark-colored and a light-colored garment possibly flip their relative appearance, in terms of which looks darker and which lighter?
Sure, if the black clothes are warmer than the white clothes, they’ll appear brighter (assuming that’s the way the scale is set up).
This can happen, but does not always happen. Is that an answer you are happy with?
To test it you could arrange a series of images with people wearing a range of different fabrics; some fabrics that look dark in visible light will look dark in IR, but others will not (and vice versa with light colours).
Near-IR and far-IR are completely different from visible light, and from each other.
Hair colour is another thing that changes, so you can’t reliably tell if someone is blond (or not) in IR.
You can get the same effect even in visible light. Have two people, one wearing red clothes and the other wearing green clothes. Illuminate them with red light, and the red clothes will look lighter. Illuminate them with green light, and the green clothes will look lighter.
You can witness this effect very easily in the numerous videos on YT where an individual is arrested and placed in the back of a patrol car. The body cams often rely solely on ambient illumination and multiple sources (streetlights, moonlight, etc.), so you will see dark hair, dark jacket, and so forth. Once the person has been placed in the back of the car, the same hair and jacket will appear to be very light in shade. This is because the camera(s) inside the vehicle are relying on IR illumination. The effects are often astonishing and you may actually question why the person with dark hair now has light hair.
I see very similar effects with my home cameras when they are operating in “night mode” with IR illumination.
It just depends on the how the features (especially clothing) reflect the IR light and how the camera’s imaging device pick it up.
Collecting together the points.
Near Infra Red, as used in security cameras is usually illuminated by the camera, and is really nothing more than a different colour. No different to red blue or green, just slightly longer in wavelength to what humans can see. In this respect, although you see a black and white image in replayed video, it isn’t showing you light and dark in the same way as a conventional black and white video would display.
It is showing you the relative brightness, and so the reflectiveness of the scene, only in the near IR “colour”. This may have little to no relationship to the colours or brightness we see normally. Just as you can’t tell how reflective a material is in red light if you are only told what the reflectiveness is with green light. It could be anything.
Near IR has no relationship to far IR, and especially nothing to do with heat imaging.
Many thanks, all! Good to know.
I vaguely recall reading decades ago that honeybees (among others) see further into near-infrared, and some flowers also have “colour” in the near-IR (reflective or absorbative of IR) that create patterns that are not obvious to us in visible light. (Typically, like other flower petal patterns, lines pointing to the center of the flower).
But yes, as others point out, what you see in near-IR depends on reflectivity of the materials. Some materials are black (absorbative) the whole range, some are reflective in IR ranges. Visible colour may not correlate.
As opposed to heat-sensing, temperature is the deciding factore in relative brightness.
I think it’s ultraviolet, not infrared, that bees see better than humans and that flowers are patterned with. Humans have better vision in long (red) wavelengths than bees do.
From early on electronic infrared imaging devices have always balanced sensitivity to produce a useful image. That prevents very hot infrared sources from washing out the entire image, differentiating from relatively close temperatures. More modern versions can adjust to the inconsistent sensitivity response from each element of a sensor array. And now color displays artificially color different temperatures as shown in @beowulff’s post. I’m a little surprised if more of that wasn’t covered by a witness providing testimony about an IR recording.
Well, depends on how hot you are.
It’s known as the PHE or Paris Hilton Effect.
DOH! Now that you mention it, yes UV - something I read almost 55 years ago.