In the Marine Corps it's taught that the 10 basic elements of camoflauge are: 1. Shape 2. Shine 3. Shadow 4. Surface 5. Sillouette 6. Sound 7. Space (Positive & Negative) 8. Color 9. Contrast 10. Movement
look up what metamaterials. All of your elements won't matter with maybe the exception of thermal. But thermal aspect can be supplied from metamaterials.
My father was a photographer in the airforce back in the 80's. He said one of his instructors was color blind & would be able to look at negatives, & instruct/recommend the student use a certain filter, adjust exposure time, adjust the shutter speed, etc, to bring out certain colors, cope for the lighting conditions, angle of the sun, etc.
Imagine how hilarious it would be if all of this was leading into TRex arms trying to find bigfoot with multi spectral imaging tech, 😆. TRex Labs has been an incredibly educational series/channel. Love it. Keep it up.
I follow a guy in Alberta, Robert Judd who caught a sasquatch in his sunglasses. He didn't see it with his eyes, but the reflection off his sunglasses shows clearly a being there. He said he had a feeling something was there though. He uses a diffuse glow filter on things he suspects are beings.
@@Ve453 I know who Lucas is and he makes one of the best holsters in the business, I just have more interests and these videos are very informative to me. There is more than guns & video games to learn & I don’t like video games.
Like and share, folks. This channel is criminally under-subscribed at this point. I'd really like them to continue making vids.Thanks for your amazing content, Isaac.
when i was a child, thermal spectroscopy was something out of a sci fi movies. snipers used to be camouflaged based on their environment and skill. I have recently seen some footage from Ukraine vs Russia conflict, where snipers were hunting snipers, and even though you can see them wearing ghillies on thermal, their bodies shine bright light or bright contrast compared to the background, making traditional camo basically useless. Thermal is scary, realll scary. Drone footage in thermal is another whole thing. there is footage out there of soldiers marching in pitch dark and they have no idea there is a tiny drone in the sky 300 ft above them, just seconds away from dropping some $5 soviet grenades, offing them all.
Tech Ingredients did a video about how to defeat the ADS (Area Denial Systems) IE microwave weapons. Basically a screen (right size mesh/material) is farraday barrier; so it'll absorb the microwaves/disperse them. I'd be interested about paints like vantablack, where you get such an absorbent surface: how does that affect lasers guided 'systems'? Do they 'follow the beam'...or look to 'target the dot' (on the target)? IF it was the latter, it'd maybe defeat that to some extent
Ultraviolet is fascinating, too. Another thing is notch filters. Active camouflage is possible using miniaturized mirrors, similar to a mirror deer blind, that are guided by computer to fool at least one known observer (solving for more is harder), the fanciest likely using DLP technology as an exterior layer.
The funny thing is one of the very first AI uses in regards to imaging was an attempt at detecting camouflaged vehicles out of various pictures. The part about that which is actually funny is they accidentally trained it to tell the difference between sunny and overcast days due to the images used as the training set. Also there's another neat kind of sensor package for imaging that has been popping up if you know what to look for, and that is THz scanning. Of course light would technically be there, but it's specifying the band of wavelengths between microwave and IR. Works similar to LIDAR, but the beam can penetrate things to some amount in order to do density mapping. It can see through some structural materials and it will also highlight stuff like any disturbed soil. So it can also spot things that are buried to some amount. So it's oddly like an X-ray of sorts, except it doesn't present the kind of hazard of X-rays or need that kind of power demand or heavy equipment to scan a large area. I figure the military uses this for some things, although it's not shown much other than a brief mention on some (futurist?) tech shows.
The first thing that came to my mind when I opened this video *was* parallax and the second was your video you posted comparing spotting scopes to long range reflex camera lenses. Therefore, the first full thought that came to my mind was a full spectrum spotting scope with a LIDAR, IR sensor, telephoto sensor, and thermal sensor. There you go,. you've got a full spectrum spotting unit that is the size of a spotting scope. It can be given to a recon team, used by PI agents, be hard mounted as a security system or mounted to some kind of ground based drone ROV. It will be the defining factor in future conflicts, the arms race of spectral awareness and advanced camouflage theory.
I've been experimenting with a similar janky camera cluster for use on a drone. I used a runcam nano for my day cam, a runcam night eagle ii for the IR and low light cam and a seek compact pro for my thermal which is running on a raspberry pi. I think I'd like to switch the Seek thermal out for an Iray unit but I'm pretty happy with the setup. Its about the size of an old gopro and while janky it absolutely works.
I think where machine learning can be very helpful is automating the process of filtering incoming data to useful ends. As shown in the video, with enough time post recording there is likely to be found multiple ways to filter image data to produce useful highlights. Sufficient compute paired with a machine learning derived filtering application likely can brute force effective filters in real time. Also likely will be interesting as subject trained models add functionality akin to switching between highlight modes(people, cars, etc). Great that Siggraph was brought up in the video. Back in high school I worked on a computer controlled paintball turret and learned lots by reading through articles and papers published by people who attended Siggraph. Always something interesting going on at Siggraph.
T rex labs is pumping out some very interesting videos lately, and from someone interested in image processing, object detection and classification this was a cool one to watch.
5:30 - want all that in one device/phone? Enter - AGM G1S - visible camera, full spectrum camera, and thermal. I just glued a 720nm filter to the full spectrum, I even had okay results(with UV illuminator) with a Kolari UV Bandpass over that camera too.
On of the Machine Learning software like IBM Maximo Visual Inspection can identify variations in frames which can then identify things moving. One you identify the things moving you can highlight them. Thus camouflage would then be useless if something was moving. Identifying something stationary would require an image without the camouflage item and one with.
Thermal reflects off of regular glass pretty well. Maybe use a piece of glass at a 45 to get the IR or visible video inline and take the thermal video reflection from the glass at 90 degrees. That should remove the parallax. (Edit) Just got to your beam-splitter reference in the video. lol
I think you should make a lens mount that has 3 mirrors in it that go to all three cameras so then you don't have the parallaxing issues since all three cameras would look through a single lens.
I have been curious about implementing a CCTV system on my property (not connected to internet of course) that can send the video back to a PC that can run some open source human detection toolbox. If any hits are registered it can send a notification and store that clip for viewing later. It would be extremely powerful if somehow I can hook up thermal video feeds to it and monitor it from a tablet over WiFi. I know some companies make this type of product package (idk about thermal) but they send all that data to their servers to crunch the detection algorithms, and who knows what they do with all that data! Probably send it right to three letter agencies or sell it to China to train their unregulated and dangerous AI models. So open source and home brewed seems like the only option
There are thermal security cameras that talk to BlueIris, and you can run it all on a PC that has no connection to the cloud. There is a lot of capability in a simple, standard, non-cloud, non-subscription, non-service, IP camera setup.
@@T.REXLabs Thanks for the info, I will definitely look into BlueIris. IP cameras using PoE seems like the cleanest/simplest way to go, and with a one-time purchase hardware license for software that can do all the motion detection alert stuff that would make this trivial to implement. I often get trapped in the loop of making sure I get the best functionality, most customizable systems for this kind of thing and get lost in the weeds when really I should just go and get something implemented. You can always upgrade software later as long as the hardware is pretty standard. I appreciate everything you guys are doing here on T-Rex Labs. I enjoy these nerdy topic discussions even more than the main channel firearm content, and I really love firearms :) keep up the good work and know it is appreciated!
@@T.REXLabsthanks for the info I will definitely look into BlueIris. Some IP cams powered via PoE seems like the simplest and cleaned way to go. Plus with a ready to go software which is a one time purchase hardware license and works disconnected from their servers sounds even better. Perhaps this will be easier than I thought to get something implemented, you can always upgrade the software later as long as your hardware setup is good. I appreciate what you guys are doing on this second channel! I love gun stuff but I love nerding out about these other topics even more 👍
An optics company needs to make a scope, red dot or prism or even binocular that immediately identifies and highlights movement in its field of view while ignoring leaves and grass blowing in the wind. That would be a game changer. Either that or we develop a movement like radar that is handheld like in the movie aliens.
Although I understand why video production tools were used for demonstration, the widest variety of tools for this exist for the GIS field, remote sensing data analysis, etc. For instance, there's a concept of feature separation metrics that analyze different combinations of multispectral channels and tell you what combination shows scene features more distinctively.
@12:48 this is a similar sensing issue to how ultrasound machines detect a heart beat. but you'd be looking for groups of movement moving in a consistent direction.
Ever read about colorblind soldiers being able to pick out Japanese soldier's camouflage? Normally colorblindness was a handicap but in this case it was an asset.
RISCV platforms have been released that work with machine intelligence of varying sorts - the big deal is that most of the projects utilize very low cost sensors and controller/processing platforms, and it's all open source - the software, hardware, and design rules/architectural language. Someone has produced (it easily crowdfunded) a *color* night vision system for about 200 USD - it uses intelligence to construct a composite image on the fly, again using multiple cheap sensors, making it look like it's simply a cloudy day as opposed to nearly completely black night.
Using IR and UV spectrum cameras you can often see camouflage items because the materials block or transmit differently at those spectrums. Some of the newer camo from military vendors are addressing the issue but more often than not stuff stands out like a sore thumb.
trex labs watch one Corridor Crew video lol, but on topic I think a leaf suit or a ghillie made from live vegetation would defeat these technics, I think thermal is still the biggest threat, yeah there are ways to hide from thermal the 2 easiest ways is hide behind glass or cover yourself in mud... yeah a movie from 1987 taught us how to defeat thermals, but also like mentioned in the video hiding amongst other hot thing like rocks could hide you but it would put you out in the open from regular human vision lol, There is also stuff like SAAB's barracuda personal camo poncho thing that supposedly hides you from thermals but for how long is yet to be seen
You have SWIR that goes to roughly 2500nm, you can see more information with it than NIR and it's not affected by things such as glass. The rest is MWIR, which is similar to thermal imaging but a tad worse.
Yeah, SWIR and MWIR see different parts of the spectrum, and they are really useful. However, cameras and optics that can see into them are not common. We will explore this part of the IR spectrum in the future.
What terrifies me is that a robotic surrogate could be equipped with all these and function semi-autonomously, it'd be able to see through walls with a swarm network of micro drones each with different sensors, able to compile a map and remove the fog of war to highlight any and all targets. Someone is going to build this, so I'd ask how to counter it.
@@T.REXLabs Depends a lot on the sensor. Agrowing offers 12MP per band on a Sony camera. You can get quite a good GSD with it. (hint, I work in defence)
Incredibly informative and enlightening (😎) video. TechnoNerd candy. I was curious that you did not mention the UV spectrum and higher frequencies. During the film/digital imagery switch, The UV photo film was suddenly made unavailable. Digital cameras that detected UV were primarily targeted to law enforcement. The camera system was export restricted and could only be sold back to the manufacturer. That was a decade or two ago. So I might be off in my current info. I speculated, at the time, that the UV spectrum was being, or had already been, developed for military and reconnaissance purposes. At all unit levels of usage. Perhaps to augment declining US superiority in Night Vision and Thermal Imaging. I got this idea from reading information about programs like Weather as a Force Multiplier: Owning the Weather in 2025 (1996) and Full Spectrum Dominance, etc.😊 Again, this is purely speculation from my reading and experience looking into UV photography.
I would be very interested to see what you think about aura/faulkner/kilner “lens filter.” ?? Also, not sure if its true but they say they can get metals(gold, silver) thin enough it becomes transparent. Wondering what that would do for a lens filter? Does it make a difference when constructing a faraday cage using different types of metals? If there is a frequency for everything what is the frequency for a “protective shield” per se. just some food for thought. Love what you do, God bless. Alex from AZ
Hate to shatter the illusion, but this has been done and dusted for a very long time - right down to automated prioritisation of threat level and automated response to those threats. There are lots of clues in the literature that look at detection, recognition and identification algorithms. As the saying goes "you can run but you can't hide." Nonetheless, good video and to be unaware is to be an early casualty.
I worked on camouflage programs both as a sniper and as a member of the defense industry building thermal camouflage products. All this makes sense to me as usable information. AI is incredibly bad. You can defeat AI with a cardboard box. A thermal camouflage with a redundant visual camouflage that is effective is not going to be defeated by this.
Everyone says "active camo" is only in bad movies and games. Wrong. I've seen actual footage where a moving human shape displaced a dust cloud climbing into a Abrahms, and an APC that could be heard, but only see the dust it kicked up. My son-in-law was a sniper for the Army for several years. He was tasked as security for a General going to an International arms show and saw someone "de-cloak" not 20 feet away. Total freak out. No sign of them to the naked eye at all. Downside? $1 million each, and the power pack only lasted 15 minutes. Some situations, that's a bargain. I used thermal in the Army (during Cold War), and it's awesome, but unless you've got some kind of notification you're not using it during the day-time.
it's all a spectrum. you have some light that is closer to visible and some that are further. Even in the thermal realm. IR =/= Thermal as if that was the case, then night vision goggles would be thermal goggles and thermal clip ons would be night vision clip ons.
@@GldenRetriever So thermal (heat) is not in the Infrared part of the spectrum? I'm guessing what you're getting at, is that The IR part of the spectrum can refer to both the IR spectrum and a specific part of the IR spectrum, and that the difference between IR and Thermal is like the difference between purple and blue in the visible spectrum. It seems really strange to me that a "color" in the IR spectrum would be called "IR". It's be like like naming "yellow" as "visible". The visible part of the visible spectrum. Kinda dumb imho. Oh well.
Technically, that area of the IR spectrum that is nearest to the visible spectrum is called Near IR (NIR), then there is Short Wave IR (SWIR), then Medium Wave IR (MWIR), and then Long Wave IR (LWIR). Night vision devices can see NIR, and thermal imagers can see LWIR. It's all on the same spectrum, just different parts.
10:29 oh yeah? What's that like? I've always wondered what having depth perception was like. (Says someone that was born with absolutely no depth perception). Lol.
Since you didn't mention motion extraction I'm compelled to mention how simple and awesome this can/could be for revealing camouflaged objects. ruclips.net/video/NSS6yAMZF78/видео.html [skip to 1m 10s for the practical application and watch through to finding the deer] I have to wonder how much more effective [digital or otherwise] night vision and such could benefit from such a simple filter method.
People have always gotten on me for just wearing black instead of having a bunch of different camouflage gear. The thing is, they were saying that while wearing Woodland back in the day and ACU at some point. I would say, camouflage in the digital age is becoming d e d dead. And we're doing all of this in the wild. Imagine trying to camouflage yourself in Compton california.
In the Marine Corps it's taught that the 10 basic elements of camoflauge are:
1. Shape
2. Shine
3. Shadow
4. Surface
5. Sillouette
6. Sound
7. Space (Positive & Negative)
8. Color
9. Contrast
10. Movement
SSSSSSSCCM. Really rolls of the tongue
look up what metamaterials. All of your elements won't matter with maybe the exception of thermal. But thermal aspect can be supplied from metamaterials.
My father was a photographer in the airforce back in the 80's. He said one of his instructors was color blind & would be able to look at negatives, & instruct/recommend the student use a certain filter, adjust exposure time, adjust the shutter speed, etc, to bring out certain colors, cope for the lighting conditions, angle of the sun, etc.
Imagine how hilarious it would be if all of this was leading into TRex arms trying to find bigfoot with multi spectral imaging tech, 😆. TRex Labs has been an incredibly educational series/channel. Love it. Keep it up.
It's nice to hear an intelligent person speaking even if you aren't familiar with the subject matter.
I follow a guy in Alberta, Robert Judd who caught a sasquatch in his sunglasses. He didn't see it with his eyes, but the reflection off his sunglasses shows clearly a being there. He said he had a feeling something was there though. He uses a diffuse glow filter on things he suspects are beings.
As a guy who uses a very expensive multispectral camera for work every day for similar applications I appreciate this.
This is more of what I wanna see from T Rex Arms!
It’s significantly better than listening to the other brother
the other brother is the face of the company and is responsible for 99% of its success. Get used to him
@@Ve453 I know who Lucas is and he makes one of the best holsters in the business, I just have more interests and these videos are very informative to me. There is more than guns & video games to learn & I don’t like video games.
Yes please more of this. This stuff is so far above the rest of Guntuber crap out there.
Like and share, folks. This channel is criminally under-subscribed at this point. I'd really like them to continue making vids.Thanks for your amazing content, Isaac.
This needs to be the top comment on every one of these videos.
when i was a child, thermal spectroscopy was something out of a sci fi movies. snipers used to be camouflaged based on their environment and skill. I have recently seen some footage from Ukraine vs Russia conflict, where snipers were hunting snipers, and even though you can see them wearing ghillies on thermal, their bodies shine bright light or bright contrast compared to the background, making traditional camo basically useless. Thermal is scary, realll scary.
Drone footage in thermal is another whole thing. there is footage out there of soldiers marching in pitch dark and they have no idea there is a tiny drone in the sky 300 ft above them, just seconds away from dropping some $5 soviet grenades, offing them all.
You guys make some of the best content on the internet. Keep it up!
Tech Ingredients did a video about how to defeat the ADS (Area Denial Systems) IE microwave weapons. Basically a screen (right size mesh/material) is farraday barrier; so it'll absorb the microwaves/disperse them.
I'd be interested about paints like vantablack, where you get such an absorbent surface: how does that affect lasers guided 'systems'? Do they 'follow the beam'...or look to 'target the dot' (on the target)? IF it was the latter, it'd maybe defeat that to some extent
Laser target designators have a hard time with materials that absorb or scatter the light, but a laser bomb that misses you by 5 feet still kills you.
The best way to defeat a laser guided weapon is to blind it with your own laser.
Ultraviolet is fascinating, too. Another thing is notch filters. Active camouflage is possible using miniaturized mirrors, similar to a mirror deer blind, that are guided by computer to fool at least one known observer (solving for more is harder), the fanciest likely using DLP technology as an exterior layer.
The funny thing is one of the very first AI uses in regards to imaging was an attempt at detecting camouflaged vehicles out of various pictures. The part about that which is actually funny is they accidentally trained it to tell the difference between sunny and overcast days due to the images used as the training set.
Also there's another neat kind of sensor package for imaging that has been popping up if you know what to look for, and that is THz scanning. Of course light would technically be there, but it's specifying the band of wavelengths between microwave and IR. Works similar to LIDAR, but the beam can penetrate things to some amount in order to do density mapping. It can see through some structural materials and it will also highlight stuff like any disturbed soil. So it can also spot things that are buried to some amount. So it's oddly like an X-ray of sorts, except it doesn't present the kind of hazard of X-rays or need that kind of power demand or heavy equipment to scan a large area. I figure the military uses this for some things, although it's not shown much other than a brief mention on some (futurist?) tech shows.
I appreciate how you are able to explain technology in such a way that is both educational and approachable. Please keep making videos!
The first thing that came to my mind when I opened this video *was* parallax and the second was your video you posted comparing spotting scopes to long range reflex camera lenses.
Therefore, the first full thought that came to my mind was a full spectrum spotting scope with a LIDAR, IR sensor, telephoto sensor, and thermal sensor.
There you go,. you've got a full spectrum spotting unit that is the size of a spotting scope. It can be given to a recon team, used by PI agents, be hard mounted as a security system or mounted to some kind of ground based drone ROV.
It will be the defining factor in future conflicts, the arms race of spectral awareness and advanced camouflage theory.
Make sure to like the video so they know that we want more stuff like this
I've been experimenting with a similar janky camera cluster for use on a drone. I used a runcam nano for my day cam, a runcam night eagle ii for the IR and low light cam and a seek compact pro for my thermal which is running on a raspberry pi. I think I'd like to switch the Seek thermal out for an Iray unit but I'm pretty happy with the setup. Its about the size of an old gopro and while janky it absolutely works.
I think where machine learning can be very helpful is automating the process of filtering incoming data to useful ends. As shown in the video, with enough time post recording there is likely to be found multiple ways to filter image data to produce useful highlights. Sufficient compute paired with a machine learning derived filtering application likely can brute force effective filters in real time. Also likely will be interesting as subject trained models add functionality akin to switching between highlight modes(people, cars, etc). Great that Siggraph was brought up in the video. Back in high school I worked on a computer controlled paintball turret and learned lots by reading through articles and papers published by people who attended Siggraph. Always something interesting going on at Siggraph.
Always. Honestly, I need to go back.
T rex labs is pumping out some very interesting videos lately, and from someone interested in image processing, object detection and classification this was a cool one to watch.
I’ve been trying to get a coyote for weeks. Thermal is a game changer
Dang! Trex broke the SFX budget on this one.
I get excited when I see you guys post videos now! Highlight of my workday!
Image capture and machine vision are amazing fields of study and this video really brings forth the excitement of exploring it, thank you!
It's all camouflaged at 144p.
This level of available technology, and knowing nation states have leaps and bounds beyond elicits anxiety.
trex taking us all to school...love it!
5:30 - want all that in one device/phone? Enter - AGM G1S - visible camera, full spectrum camera, and thermal. I just glued a 720nm filter to the full spectrum, I even had okay results(with UV illuminator) with a Kolari UV Bandpass over that camera too.
On of the Machine Learning software like IBM Maximo Visual Inspection can identify variations in frames which can then identify things moving. One you identify the things moving you can highlight them. Thus camouflage would then be useless if something was moving. Identifying something stationary would require an image without the camouflage item and one with.
Thanks for posting
Isaac is a wizard with the imagery and software! T.Rex arms robotic holster with CPU and sensor suite coming soon?
Thermal reflects off of regular glass pretty well. Maybe use a piece of glass at a 45 to get the IR or visible video inline and take the thermal video reflection from the glass at 90 degrees. That should remove the parallax.
(Edit) Just got to your beam-splitter reference in the video. lol
I love the weeds...unless they are in my garden. Keep them coming.
You completely brushed over the most useful technique for discovering camouflaged persons. It is completely pattern based...
I think you should make a lens mount that has 3 mirrors in it that go to all three cameras so then you don't have the parallaxing issues since all three cameras would look through a single lens.
It is really interesting to see the overlap the concepts of this video with the astronomy world
Instead of fighting parallax, have you considered using it to find depth in the way 3D cameras and you eyes work?
It requires some more processing power but it's a great thought!
But can advanced technology and AI see why kids love the taste of Cinnamon Toast Crunch
Or catch that Trixy rabbit?
Or tell me why apple jacks don't taste like apples?
or jacks
I have been curious about implementing a CCTV system on my property (not connected to internet of course) that can send the video back to a PC that can run some open source human detection toolbox. If any hits are registered it can send a notification and store that clip for viewing later. It would be extremely powerful if somehow I can hook up thermal video feeds to it and monitor it from a tablet over WiFi. I know some companies make this type of product package (idk about thermal) but they send all that data to their servers to crunch the detection algorithms, and who knows what they do with all that data! Probably send it right to three letter agencies or sell it to China to train their unregulated and dangerous AI models. So open source and home brewed seems like the only option
There are thermal security cameras that talk to BlueIris, and you can run it all on a PC that has no connection to the cloud. There is a lot of capability in a simple, standard, non-cloud, non-subscription, non-service, IP camera setup.
@@T.REXLabs Thanks for the info, I will definitely look into BlueIris. IP cameras using PoE seems like the cleanest/simplest way to go, and with a one-time purchase hardware license for software that can do all the motion detection alert stuff that would make this trivial to implement. I often get trapped in the loop of making sure I get the best functionality, most customizable systems for this kind of thing and get lost in the weeds when really I should just go and get something implemented. You can always upgrade software later as long as the hardware is pretty standard. I appreciate everything you guys are doing here on T-Rex Labs. I enjoy these nerdy topic discussions even more than the main channel firearm content, and I really love firearms :) keep up the good work and know it is appreciated!
@@T.REXLabsthanks for the info I will definitely look into BlueIris. Some IP cams powered via PoE seems like the simplest and cleaned way to go. Plus with a ready to go software which is a one time purchase hardware license and works disconnected from their servers sounds even better. Perhaps this will be easier than I thought to get something implemented, you can always upgrade the software later as long as your hardware setup is good. I appreciate what you guys are doing on this second channel! I love gun stuff but I love nerding out about these other topics even more 👍
Thank you for mentioning ACM and SIGGRAPH. :)
An optics company needs to make a scope, red dot or prism or even binocular that immediately identifies and highlights movement in its field of view while ignoring leaves and grass blowing in the wind. That would be a game changer. Either that or we develop a movement like radar that is handheld like in the movie aliens.
Although I understand why video production tools were used for demonstration, the widest variety of tools for this exist for the GIS field, remote sensing data analysis, etc. For instance, there's a concept of feature separation metrics that analyze different combinations of multispectral channels and tell you what combination shows scene features more distinctively.
Fantastic content!! Keep it up. Thank you Isaac for sharing your knowledge.
You should look into Starvis2 sensors as well, they're... fascinating.
@12:48 this is a similar sensing issue to how ultrasound machines detect a heart beat. but you'd be looking for groups of movement moving in a consistent direction.
That was awesome. The creativity amazes me.
Ever read about colorblind soldiers being able to pick out Japanese soldier's camouflage? Normally colorblindness was a handicap but in this case it was an asset.
Really cool. I think camo is an understudied area. You brought up some really cool ideas. Thank you.
RF/EMF snifer and drone the guy from several miles away, walking around with all the imagining gear.
Basically what Predator did in the 2nd movie when they thought they could hide in the meat house. Love that movie.
RISCV platforms have been released that work with machine intelligence of varying sorts - the big deal is that most of the projects utilize very low cost sensors and controller/processing platforms, and it's all open source - the software, hardware, and design rules/architectural language.
Someone has produced (it easily crowdfunded) a *color* night vision system for about 200 USD - it uses intelligence to construct a composite image on the fly, again using multiple cheap sensors, making it look like it's simply a cloudy day as opposed to nearly completely black night.
Interesting, do you recall the name of that project?
Another vid! Let’s go!
Awesome possum!!! That turkey camouflaged!!! What the heck???
14:18 You don’t find out about it until it’s used a few times lmao.
Using IR and UV spectrum cameras you can often see camouflage items because the materials block or transmit differently at those spectrums. Some of the newer camo from military vendors are addressing the issue but more often than not stuff stands out like a sore thumb.
trex labs watch one Corridor Crew video lol, but on topic I think a leaf suit or a ghillie made from live vegetation would defeat these technics, I think thermal is still the biggest threat, yeah there are ways to hide from thermal the 2 easiest ways is hide behind glass or cover yourself in mud... yeah a movie from 1987 taught us how to defeat thermals, but also like mentioned in the video hiding amongst other hot thing like rocks could hide you but it would put you out in the open from regular human vision lol, There is also stuff like SAAB's barracuda personal camo poncho thing that supposedly hides you from thermals but for how long is yet to be seen
This is one video that is way out of my wheelhouse, but it was still interesting
I think most camouflages can be seen in the UV-A and B spectrums. Albeit dark because of the scattering.
Could you do a video on how to defeat cameras that can detect concealed weapons?
Thanks Isaac!
Edit: I guessed Avatar (blue people) had the best compositing lol gotta love these fun facts
4:30 What does the 1500 - 8000 nm spectrum look like? There's a giant gap there.
You have SWIR that goes to roughly 2500nm, you can see more information with it than NIR and it's not affected by things such as glass. The rest is MWIR, which is similar to thermal imaging but a tad worse.
Yeah, SWIR and MWIR see different parts of the spectrum, and they are really useful. However, cameras and optics that can see into them are not common. We will explore this part of the IR spectrum in the future.
What terrifies me is that a robotic surrogate could be equipped with all these and function semi-autonomously, it'd be able to see through walls with a swarm network of micro drones each with different sensors, able to compile a map and remove the fog of war to highlight any and all targets. Someone is going to build this, so I'd ask how to counter it.
Holosun's new thermal nightvision sight thing they have is super powerful in this regard, and only for 1600 bucks
Unfortunately, their thermal lense is not able to pick up heat from very far away.
@@kerbalairforce8802 it's within average shooting range, it's defiantly decent. Not great for recon, but good for general use I think
You can use multispectral to detect in real time by overlaying the colors.
You can, but it starts to get really noisy. We should have showed that.
@@T.REXLabs Depends a lot on the sensor. Agrowing offers 12MP per band on a Sony camera. You can get quite a good GSD with it. (hint, I work in defence)
Fascinating stuff. Well done
Please make some videos reviewing movies you love and what sticks out to you from those films
Incredibly informative and enlightening (😎) video. TechnoNerd candy.
I was curious that you did not mention the UV spectrum and higher frequencies.
During the film/digital imagery switch, The UV photo film was suddenly made unavailable. Digital cameras that detected UV were primarily targeted to law enforcement. The camera system was export restricted and could only be sold back to the manufacturer. That was a decade or two ago. So I might be off in my current info.
I speculated, at the time, that the UV spectrum was being, or had already been, developed for military and reconnaissance purposes. At all unit levels of usage. Perhaps to augment declining US superiority in Night Vision and Thermal Imaging.
I got this idea from reading information about programs like Weather as a Force Multiplier: Owning the Weather in 2025 (1996) and Full Spectrum Dominance, etc.😊
Again, this is purely speculation from my reading and experience looking into UV photography.
I would be very interested to see what you think about aura/faulkner/kilner “lens filter.” ??
Also, not sure if its true but they say they can get metals(gold, silver) thin enough it becomes transparent. Wondering what that would do for a lens filter?
Does it make a difference when constructing a faraday cage using different types of metals?
If there is a frequency for everything what is the frequency for a “protective shield” per se.
just some food for thought. Love what you do, God bless.
Alex from AZ
You don't HAVE to subscribe to after effects. That's assuming you have a handy eye patch laying around on your ship.
Hearing some audio issues in the shots where you are outside.
I was wondering about that myself.
Awesome, Cheers
Hate to shatter the illusion, but this has been done and dusted for a very long time - right down to automated prioritisation of threat level and automated response to those threats. There are lots of clues in the literature that look at detection, recognition and identification algorithms. As the saying goes "you can run but you can't hide." Nonetheless, good video and to be unaware is to be an early casualty.
This applies to shelter camouflage as well.
I worked on camouflage programs both as a sniper and as a member of the defense industry building thermal camouflage products. All this makes sense to me as usable information. AI is incredibly bad. You can defeat AI with a cardboard box. A thermal camouflage with a redundant visual camouflage that is effective is not going to be defeated by this.
Incredible video mate
your thumbnail forgot the U in camouflage
It’s something different, 90% of the views wouldn’t catch that. Just like the video topic😂
whoops
@@T.REXLabs You put it there, it was just well hidden.
Everyone says "active camo" is only in bad movies and games. Wrong. I've seen actual footage where a moving human shape displaced a dust cloud climbing into a Abrahms, and an APC that could be heard, but only see the dust it kicked up. My son-in-law was a sniper for the Army for several years. He was tasked as security for a General going to an International arms show and saw someone "de-cloak" not 20 feet away. Total freak out. No sign of them to the naked eye at all. Downside? $1 million each, and the power pack only lasted 15 minutes. Some situations, that's a bargain. I used thermal in the Army (during Cold War), and it's awesome, but unless you've got some kind of notification you're not using it during the day-time.
Have you tried automating these techniques?
Epic
keep it up guys.
This is cool stuff. Love it!
Funny timing with the brass facts video
Seding love💚"Sir"
Good video!
This is how you analyze intelligence photos
the absence of brent0331 in this episode is unforgivable
Wait a second... aren't I.R. and Thermal two names for the same thing? Thermal imaging lies on the InfraRed part of the light spectrum.
it's all a spectrum. you have some light that is closer to visible and some that are further. Even in the thermal realm.
IR =/= Thermal as if that was the case, then night vision goggles would be thermal goggles and thermal clip ons would be night vision clip ons.
@@GldenRetriever So thermal (heat) is not in the Infrared part of the spectrum? I'm guessing what you're getting at, is that The IR part of the spectrum can refer to both the IR spectrum and a specific part of the IR spectrum, and that the difference between IR and Thermal is like the difference between purple and blue in the visible spectrum. It seems really strange to me that a "color" in the IR spectrum would be called "IR". It's be like like naming "yellow" as "visible". The visible part of the visible spectrum. Kinda dumb imho. Oh well.
Technically, that area of the IR spectrum that is nearest to the visible spectrum is called Near IR (NIR), then there is Short Wave IR (SWIR), then Medium Wave IR (MWIR), and then Long Wave IR (LWIR). Night vision devices can see NIR, and thermal imagers can see LWIR. It's all on the same spectrum, just different parts.
@@T.REXLabsbe interesting to throw ultra violet in with the rest.
What thermal setup was that?
Was that KSP used for the satellite clip?
QGIS.
So embrace tradition reject modernity? Go full Arnold in mud packs and a explody bow!
Arent Modern "Seekers" on Rockets and such searchin and "lookin" in the UltraViolette Spectrum too ?
Broad spectrum cameras are expensive, and not in common use for disposable weapons.
I'd be more concerned about synthetic apature radar.
great video but active camouflage is no where even ready yet,
link to your sigraph paper?
You should reach us for these topics
Nazgol tracking hobbits in the woods...
Yo how much did that jank camera setup cost? Looks like several thousand dollars
Well Golly, answered at @25:30
Does this work with the car keys?
10:29 oh yeah? What's that like? I've always wondered what having depth perception was like. (Says someone that was born with absolutely no depth perception). Lol.
Thank you for not making this a political program
Project 2025 will put multispectral imaging technology into the hands of private American citizens.
Unlike the other brother…
@@Thee_Sinner you must not be very familiar with me. If you've ever listened to the podcast, you'd know that isn't true.
@@isaacbotkintrexsounds like you need to get more political
@@isaacbotkintrex I know enough :)
you didn't smear mud all over yourself to blind the predator. I'm dissapoint.
Since you didn't mention motion extraction I'm compelled to mention how simple and awesome this can/could be for revealing camouflaged objects. ruclips.net/video/NSS6yAMZF78/видео.html [skip to 1m 10s for the practical application and watch through to finding the deer]
I have to wonder how much more effective [digital or otherwise] night vision and such could benefit from such a simple filter method.
Realtime application if not well developed: ruclips.net/video/6KYUE2oy5Bo/видео.html
@Corridor
People have always gotten on me for just wearing black instead of having a bunch of different camouflage gear. The thing is, they were saying that while wearing Woodland back in the day and ACU at some point. I would say, camouflage in the digital age is becoming d e d dead. And we're doing all of this in the wild. Imagine trying to camouflage yourself in Compton california.