Also leads to some interesting possibilities - given that this occured in November in Minneapolis, there seems like there would be opportunities to manipulate the infrared camera. Off the top of my head - extra hot point heat sources could "blow out" the image, or insulated umbrellas/barriers could obfuscate where or how many protesters are present.
Could anyone with infrared camera experience weigh in?
While I haven't used the equipment here, I own a number of consumer- and low-end professional-grade thermal imagers. Resolution and NETD differ, but I expect the overall behaviors are similar.
They're remarkably good at not letting point sources blow out the image. That's a very important performance parameter, since often you're not concerned with the hottest point, but the parts immediately adjacent, so you can see how heat is flowing through the device being inspected. A lot of investment and IP goes into this -- since the hot spot _will_ heat the imaging array, and on older imagers it would cause streaking if moved, modern ones do extensive thermal modeling of the array itself to try to cancel out this effect, all baked into FPGAs so it can be done in real-time. They're very good at it.
A very naïve implementation may have auto-scaling for the color palette, in which case a hot-spot would skew the palette and make cooler stuff hard to distinguish, but that would be a total amateur mistake. Setting the palette to highlight human skin temperature is like step 1 of any such operation.
Insulated umbrellas, or just big aluminized-mylar space-blankets, would definitely play havoc with the image. An AI people-counter trained on the image would probably lose its marbles, but a human looking at it would quickly understand what's going on and make some educated guesses. It's very hard to hide your thermal signature for very long, which is why thermal imaging is so important in warfighting.
I'm shocked how terrible quality the audio comms are.