Color Night Vision!

https://www.x20.org/color-night-vision/

I have no idea how this is possible, but it's very cool (assuming it's real) - especially the FLIR overlay.

Why would it not be possible?

Older night vision (low light not, infra-red) used to take all the light and amplify it so you can see it. They did this because the sensors used to be inherently noisy, so taking all the light and just making a grayscale image resulted in better images.

Today we have sensors that doesn't have such a large inherent noise so you can take red/green/blue light seperately and make up a colour image.

Simplified it works like this:

For each image, each pixel gets a value: p = (R + G + B) + N

where RGB is Red, Green, Blue photon counts and N is noise in the sensor. N used to be quite large. Nowadays it is much smaller so you can have each in their own channel and the noise doesn't overwhelm the photon counts in low light. Then afterwards you can make up a colour image from the R,G and B images.

The low noise is the amazing thing and I'm also really wondering how it's possible. Stating that there's nothing amazing here as it just has low noise bause the the year in the calendar is 2017 is like saying that there's nothing special about Einstein, he's only really smart.

Maybe it has a Peltier cooler and CMOS chip? You can see outside in the dark (no colour) on a dark moonless but clear night and your eyes have a quantum efficieny (% of photons caught) of 1% or so. CCDs and CMOS upwards of 90%. Add a big lense, a physically large sensor with physically large pixels (catches more photons) and you have quite a few photons to work with.

Source: I'm an astronomer.

Light amplification in NVG's does not amplify the actual photons coming in from a source. Those incoming photons are just used to shoot a large number of electrons at a phosphor plate. The color of the NVG's depends on the phosphor. Most are green, but the NVG's I fly with now are white. But at any rate you are not seeing the photons from the source, so it's impossible to see color with the way they are built today.

Effectively they have a low noise detector that does the same as a plain NVG but for 3 different color bands. Imagine you had 3 NVGs, one with a red filter, one with green and one with blue. You take the image from each and make a color image and show it to the wearer.

Maybe it has a Peltier cooler and CMOS chip? You can see outside in the dark (no colour) on a dark moonless but clear night and your eyes have a quantum efficieny (% of photons caught) of 1% or so. CCDs and CMOS upwards of 90%. Add a big lense, a physically large sensor with physically large pixels (catches more photons) and you have quite a few photons to work with.

 

Source: I'm an astronomer.

I think this makes the most sense. It's pretty absurd how clean the signal is from an actively cooled high quality sensor. It's just, like, really expensive though.

Effectively they have a low noise detector that does the same as a plain NVG but for 3 different color bands. Imagine you had 3 NVGs, one with a red filter, one with green and one with blue. You take the image from each and make a color image and show it to the wearer.

Yeah I get what you're saying. The problem I'm seeing is that the color you see has nothing to do with the colors that come into the tube. What color you see depends purely on the composition of the phosphor screen. I suppose you could have three different phosphor screens for each of the three filters, but I don't see how you're going to combine those three images in a way that you could put it onto a human head.

the op link isnt an nvg unit but since we're talking about it, it seems reasonable to me that in the future nvgs will move away from phosphor plates to cmos outputting into oleds.

According to the advertisement materials it uses something called Broad Spectrum Thin Film Array. Reading through the materials gave me the impression it's a multi layer CMOS sensor with very little empty space between pixels and some kind of nanoparticle film improving it's sensitivity. It's also paired with very capable image processors but it's unclear if it's any better than state of the art image processors from Canon or Sony for example but certainly not worse. Essentially everyday camera sensor technology doped up to the absolute edge of performance that is possible with current knowledge. I think this patent could be just the thing, at least it descibes an imaging sensor with the same kind of performance as we can see in the videos.

https://www.google.com/patents/US20050104089